Abstract
Superficial siderosis is a rare disorder characterised by the deposition of haemosiderin on the surface of the central nervous system. Cognitive dysfunction has sporadically been reported in relation with superficial siderosis. We present a 61-year-old man with cognitive dysfunction in the presence of the typical radiological image of temporal and cerebellar superficial siderosis, most likely due to pseudomeningocoele 14 years after resection of a meningioma at the cervicothoracic junction. Xantochromia was present on cerebrospinal fluid investigation and a source of bleeding was seen during surgical exploration. Despite surgical treatment of the suspected bleeding source, the patient deteriorated and neuropsychological examination 1 year after surgery showed progression of cognitive dysfunction to dementia. It is likely that in the absence of other typical symptoms such as cerebellar ataxia and hearing loss, the cognitive dysfunction was not related to the superficial siderosis.
Keywords: memory disorders, neurosurgery, spinal cord
Background
Superficial siderosis (SS) is a rare disorder characterised by the deposition of haemosiderin on the surface of the central nervous system (CNS), most commonly the cortex of the cerebellum, the vestibulocochlear nerve and the brain stem. Progressive ataxia, sensorineural hearing loss and corticospinal tract signs have been reported. Cognitive dysfunction has sporadically been reported in relation with SS.1 2 We present a case with cognitive dysfunction in the presence of this rare disorder, caused by persistent erythrocyte leakage in the cerebrospinal fluid 14 years after resection of a spinal meningioma at the cervicothoracic junction. Neurosurgical re-exploration confirmed the bleeding source, intraoperative images of which are shown.
Case presentation
A 61-year-old man presented with progressive cognitive dysfunction since 2 years. He received 10 years of education and finished a lower vocational education as an electronic technician. He had worked as an electronic technician and a military driving instructor, and when he resigned from the military, he started as a maintenance technician. He reported increasing problems in performing his current job as a maintenance technician and problems with concentration, short-term memory and orientation. He suffered from tinnitus for several years. He reported hearing problems, primarily with following conversations in larger groups, but on evaluation, there was no objective hearing loss. His medical history included resection (Simpson grade I) of a spinal meningioma at C7-Th1 level 14 years before. He used statin, no other medication. There was no history of alcohol or substance abuse. His family history was negative for psychiatric diseases or early onset dementia. Neurological examination showed no abnormalities, in particular, no ataxia and no pyramidal signs. Score on mini-mental state examination was 23/30 on first presentation. Neuropsychological assessment (NPA) including tests on processing speed, attention, memory, visuoconstructive ability, executive function and language confirmed mild cognitive impairment (MCI) in multiple domains.
Investigations
Routine laboratory examination (including blood count, sedimentation rate, liver and renal function tests, glucose, electrolytes, thyroid function, vitamin B12 and screening on Borrelia and syphilis) was normal. MRI of the brain showed medial temporal atrophy (MTA) score 2 (figure 1), substantial periventricular and subcortical white matter hyperintensities (Fazekas 2) as well as susceptibility artefacts on pial and ependymal surfaces of primarily the cerebellum and temporal lobes (figure 2A, B). Magnetic resonance angiography showed no aneurysms or other vascular abnormalities. MRI of the cervical spine showed a right-sided caudal decompression of the C7 lamina and a right-sided haemilaminectomy of Th1 at the site of meningioma resection and a cystic lesion directly posterior to the dura (figure 2C, D). There appeared to be a connection between the cyst and the cerebrospinal fluid (CSF) at the Th1 right-sided neuroforamen. Based on the available literature, it was postulated that the cyst could be a source of persistent erythrocyte leakage, resulting in SS. An atraumatic lumbar puncture was performed to evaluate the presence of red blood cells. CSF analysis showed the presence of xantochromia (bilirubin 0.5 mmol/L) and elevated red blood cells (3000×106/L), without pleocytosis (white blood cell in CSF 7000×109/L). Tau biomarkers and biomarkers for prion disease were not determined.
Figure 1.

Coronal T1-weighed MRI image of the brain depicting medial temporal lobe with a medial temporal atrophy score of 2.
Figure 2.
(A, B) Axial gradient echo MRI images showing areas of hypointensity, primarily located at the cerebellar surface and the mesial temporal lobes. (C, D) Sagittal and axial T2-weighed MRI images showing the cyst. On the axial image, a suspected connection between the cyst and the dural sleeve of the right Th1 nerve root can be appreciated (arrow).
Differential diagnosis
The radiological diagnosis of SS on MRI of the brain, in the presence of a suspected source of erythrocyte leakage on MRI of the cervical spine, and xantochromia in CSF obtained by lumbar puncture make a causal relationship between the SS and the cyst plausible. Cognitive dysfunction has sporadically been described in patients with SS, without a specific profile.
The profile of the NPA did not match a specific profile, such as can be found in some conditions (eg, Alzheimer’s disease). In our case, we considered vascular cognitive impairment as well as cognitive dysfunction due to SS.
Treatment
As SS could not be ruled out as the cause for the cognitive dysfunction and the underlying erythrocyte leakage was potentially treatable, exploration of the cyst was offered to the patient. The goal of the surgery was to stop further progression of the cognitive dysfunction. NPA was repeated preoperatively (1 year after the initial NPA) to serve as a baseline measurement before treatment. This showed progression of cognitive dysfunction from MCI to dementia with further impairment in orientation and attention, memory, language skills, (complex) visuoconstruction and executive functions.
Re-exploration was performed through the previous midline incision. The cyst was opened, which on examination was communicating with the dural sac through a defect in the dural sleeve of the Th1 root, which was clearly visible (figure 3). Interestingly, the caudal end of the C7 lamina was involved in the cyst wall. The cyst wall at this location was erythematous with yellowish discolourations suggesting haemosiderin depositions around it as well as a small blood clot in the roof of the cyst (figure 3). The connection of the cyst to the dural sleeve was closed using sutures and fibrin sealant. The cranial side of the cyst, where the wall was continuous with the previously partially resected lamina of C7, was coagulated. The lateral walls of the cyst were vitalised and then sutured together, after which fibrin sealant was applied. The wound was meticulously closed in layers.
Figure 3.
Intraoperative images. In the left image, the opened cyst is visible with at the cranial aspect, continuous with the caudal aspect of the previously partially resected lamina of C7, a blood clot with surrounding yellowish discolouration due to suspected haemosiderin deposits (asterisk). In the right image, the connection of the cyst to the dural sleeve is visible, with the right Th1 nerve root visible through the opening (arrow). In this image, it is visible that the cyst wall adherent to the caudal part of lamina C7 was coagulated.
Outcome and follow-up
Despite surgical treatment, the patient deteriorated. NPA was repeated 1 year after surgery and showed further progression of dementia. This decline was present in all domains: orientation and attention, memory, language skills, visuoconstruction and executive functions. Word-finding difficulties were noted when speaking with the patient. The patient was increasingly dependent on his wife. Repeat MRI of the cervical spine 1 year after surgery showed a slight decrease in the fluid-filled cavity, the previously present connection with the subarachnoid space was no longer present. Repeat MRI brain 2 years after surgery showed a similar dissemination of the susceptibility artefacts, with a slight increase in subcortical white matter hyperintensities and progression of MTA score 3.
Three years after surgery, he was placed in a care home.
Discussion
A suspected bleeding source cannot always be identified in SS of the CNS, but previous neurosurgical procedures have been reported as one of the predisposing factors.3–6 In a series of 30 patients, Kumar et al reported the presence of a pseudomeningocoele or fluid-filled collection in 14 patients, often at a spinal location, similar to our case.6
In case of a pseudomeningocoele as the suspected cause for SS, surgery can be considered.7–12 Progression of symptoms may be halted, although in cases of a long history of SS further progression despite dural closure has been reported.7 A previous case report reported an increased vascularity of the dura on re-exploration of a postsurgery cyst.11 In that case, histopathology showed paucicellular hyalinised collagen tissue with multifocal reactive vascular proliferation. In our case, there also was an increased vascularity of the cyst wall at the site where it was adherent to the caudal end of the partially resected lamina of C7. The yellowish discolouration suggests haemosiderin deposits due to chronic or recurrent haemorrhaging at this location. This is strengthened by the fact that on opening the cyst, the blood clot was already present at that location, while care had been taken to not let blood enter the cyst cavity. A fairly similar bleeding source was described previously, where the bone marrow vasculature of the remaining vertebral arch following cervical laminectomy for the removal of a spinal schwannoma was identified as the source.13
SS should not be confused with cortical SS, which has a different distribution over the supratentorial convexity and is mostly caused by amyloid angiopathy, a condition that is clinically characterised by intracerebral haemorrhage and cognitive decline.14 Our patient did not have extensive cortical presence of SS. However, the mesiotemporal SS could possibly account for the cognitive dysfunction in our case. Dementia was described as a clinical finding in SS in 3 out of 23 patients in a review of the literature, while in their own retrospective MRI case series, it was reported in 1 out of 13 patients.5
To date, no other cases of surgically treated patients with cognitive dysfunction due to SS with a neuropsychological follow-up have been reported. It is, therefore, not known if resolution of a suspected bleeding source can stop the progression of the cognitive dysfunction. Given the fact that there was a further progression of the cognitive dysfunction, in combination with the findings on repeat MRI brain, it is highly likely that the patient in our case has a neurodegenerative or vascular cause of his dementia. Noteworthy is the fact that, contrary to the patient in this case report, all three patients who were previously described with dementia as a symptom of SS also had cerebellar ataxia and hearing loss,5 and cognitive dysfunction should probably be considered a late-stage symptom.1 A radiological diagnosis of SS might, therefore, be a chance finding in patients with isolated cognitive dysfunction.
Learning points.
Superficial siderosis (SS) is a rare disorder characterised by the deposition of haemosiderin on the surface of the central nervous system.
Cognitive dysfunction has sporadically been reported in relation with SS.
Previous neurosurgical procedures have been reported as one of the predisposing factors for SS.
A pseudomeningocoele has been reported in up to 47% of patients with SS.
Absence of hearing loss and cerebellar ataxia might be a sign that in patients with cognitive dysfunction, a radiological diagnosis of SS is a chance finding.
Acknowledgments
We would like to thank WM Pennarts, MSc, Clinical Neuropsychologist, for her valuable comments and revision of the manuscript.
Footnotes
Contributors: TFMB wrote the manuscript with support from JvA and JS.
Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests: None declared.
Patient consent for publication: Next of kin consent obtained.
Provenance and peer review: Not commissioned; externally peer reviewed.
References
- 1.van Harskamp NJ, Rudge P, Cipolotti L. Cognitive and social impairments in patients with superficial siderosis. Brain 2005;128:1082–92. 10.1093/brain/awh487 [DOI] [PubMed] [Google Scholar]
- 2.Uttner I, Tumani H, Arnim Cvon, et al. Cognitive impairment in superficial siderosis of the central nervous system: a case report. Cerebellum 2009;8:61–3. 10.1007/s12311-008-0073-5 [DOI] [PubMed] [Google Scholar]
- 3.Kumar R, Jacob JT, Welker KM, et al. Superficial siderosis of the central nervous system associated with incomplete dural closure following posterior fossa surgery: report of 3 cases. J Neurosurg 2015;123:1326–30. 10.3171/2014.12.JNS141920 [DOI] [PubMed] [Google Scholar]
- 4.McCarron MO, Flynn PA, Owens C, et al. Superficial siderosis of the central nervous system many years after neurosurgical procedures. J Neurol Neurosurg Psychiatry 2003;74:1326–8. 10.1136/jnnp.74.9.1326 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Offenbacher H, Fazekas F, Schmidt R, et al. Superficial siderosis of the central nervous system: MRI findings and clinical significance. Neuroradiology 1996;38 Suppl 1:S51–6. 10.1007/BF02278119 [DOI] [PubMed] [Google Scholar]
- 6.Kumar N, Cohen-Gadol AA, Wright RA, et al. Superficial siderosis. Neurology 2006;66:1144–52. 10.1212/01.wnl.0000208510.76323.5b [DOI] [PubMed] [Google Scholar]
- 7.Egawa S, Yoshii T, Sakaki K, et al. Dural closure for the treatment of superficial siderosis. J Neurosurg Spine 2013;18:388–93. 10.3171/2013.1.SPINE12649 [DOI] [PubMed] [Google Scholar]
- 8.Kumar N. Superficial siderosis: associations and therapeutic implications. Arch Neurol 2007;64:491–6. 10.1001/archneur.64.4.491 [DOI] [PubMed] [Google Scholar]
- 9.Kumar N, Bledsoe JM, Davis DH. Intracranial fluid filled collection and superficial siderosis. J Neurol Neurosurg Psychiatry 2007;78:652–3. 10.1136/jnnp.2006.108225 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kumar N, Lane JI, Piepgras DG. Superficial siderosis: sealing the defect. Neurology 2009;72:671–3. 10.1212/01.wnl.0000342457.22536.af [DOI] [PubMed] [Google Scholar]
- 11.Lekgabe E, Kavar B. Progression and management of superficial siderosis. J Clin Neurosci 2012;19:906–8. 10.1016/j.jocn.2011.08.029 [DOI] [PubMed] [Google Scholar]
- 12.Shih P, Yang BP, Batjer HH, et al. Surgical management of superficial siderosis. Spine J 2009;9:e16–19. 10.1016/j.spinee.2009.03.004 [DOI] [PubMed] [Google Scholar]
- 13.Yokosuka J, Takai K, Komori T, et al. Superficial siderosis: bleeding from the bone marrow after laminectomy for spinal tumor removal. J Neurosurg Spine 2014;21:905–8. 10.3171/2014.8.SPINE13328 [DOI] [PubMed] [Google Scholar]
- 14.Charidimou A, Linn J, Vernooij MW, et al. Cortical superficial siderosis: detection and clinical significance in cerebral amyloid angiopathy and related conditions. Brain 2015;138:2126–39. 10.1093/brain/awv162 [DOI] [PubMed] [Google Scholar]


