ECD is a rare non-Langerhans cell histiocytosis that causes a multisystemic disease that can compromise long bones with painful osteoclastic lesions, and can also involve retroperitoneum, cardiovascular, pulmonary or orbits and cause periaortitis, pericarditis, retroperitoneal fibrosis, pulmonary infiltrates or xanthelasmas [1, 8]. Its CNS involvement is relatively common, around 50%, and can mimic various intra and extra-axial lesions generating a widespread of neurological manifestations and/or complications such as diabetes insipidus and hydrocephalus [6, 9]. Its diagnosis is achieved by pathological examination, imaging findings and genetic studies.
On the other hand, neurosarcoidosis is an uncommon manifestation of sarcoidosis seen in about 5-35% of patients with sarcoidosis, occasionally presented in isolation without signs of systemic illness. It’s CNS compromise includes meningeal, cranial nerves, hypothalamus, spinal cord or peripheral nerve involvement showing granulomatous infiltrates in the pathology . Its diagnosis requires all clinical histological, imaging, and laboratory test support, and given its broad clinical spectrum, it is considered one of the great imitators, masquerading itself as other diseases becoming a challenging diagnosis to come by, as well as a powerful confusing factor as seen in our case, initially considered to be a posterior fossa tumour.
The MRI showed that the posterior fossa lesion was isointense in those T1 and T2-weighted images with high contrast enhancement, and the orbit lesions were hypointense in the same sequences, this finding it’s not specific for neither of the conditions mentioned above. Both present with a meningeal enhancement and hyperintense lesions on the FLAIR sequences. For neurosarcoidosis the imaging findings can be classified according to Christoforidis et al, 1999, in six categories, being: (1) dural thickening or mass, (2) leptomeningeal involvement, (3) enhancing brain parenchymal lesion, (4) non enhancing brain parenchymal lesion, (5) cranial nerve involvement and (6) spinal cord and root involvement . In our case the patient did not fit just one of the categories mentioned above given that her lesion had both dural thickening or mass and cranial nerve involvement.
No such classification exists for ECD but given the cases reported in literature it could be divided according to the location of the lesions: (1) meningeal involvement with thickening and/or mass, (2) hypothalamic-pituitary axis mass, (3) supratentorial brain parenchymal lesion excluding hypothalamic-pituitary axis, (4) infratentorial parenchymal mass lesions and, (5) cerebrovascular involvement with peri-arterial infiltration. Taking this into account, our patient showed meningeal involvement with a mass lesion, that is seen in 23% of ECD cases, and cerebrovascular involvement with orbital lesions . The concomitant presentation with intraconal orbital lesions that are seen hypointense in the T2-weighted sequences are indicative of ECD and should expedite diagnosis as does the coexistence of neurologic symptoms that are not attributed to the intracranial lesion such as diabetes insipidus or cerebellar symptoms .
Initial testing did not include CSF analysis given that the patient presented itself with a posterior fossa tumour and the risk of the performing a lumbar puncture was high, therefore omitted. Pathology studies of both diseases show inflammatory changes within the tissue with avid lymphocytic infiltration and granulomatous lesions formation. Differentiation between both of them can be challenging and not even immunohistochemistry can distinguish them. For ECD, granulomas present with positive CD68 and negative CD1a, 20% of cases can present positive S100 which correlates with what’s seen in the patients’ pathology, but it also presents lymphocytic infiltrates of CD3 T-lymphocytes and B-lymphocytes positive for CD20 that is present in neurosarcoidosis [2, 5]. The lack of Birbeck granules in the electronic microscopy could have helped provide an early diagnosis of ECD, but it was not analysed in our case due to the lack of this resource in our institution .
At this stage, the diagnosis of ECD was obscured by the possibility of neurosarcoidosis given that both diseases can present similar clinical manifestations, imaging findings on MRI and even undifferentiated histopathological infiltrates. Other differential diagnoses that could be taken into consideration are Langerhans cell histiocytosis, Rosai-Dorfman disease, juvenile xanthogranuloma, multiple sclerosis and granulomatous infectious diseases like tuberculosis, toxoplasmosis and fungal infections. Since tuberculosis was considered in our case a Chest CT was obtained giving the diagnostic possibility of ECD. It did not only show the presence of sclerotic lesions but also of periaortitis, pulmonary infiltrates and pericardial and pleural effusion. The confirmation was made by bone scintigraphy in which bilateral symmetric metadiaphyseal medullary sclerosis with epiphyseal sparring and intense uptake of the radiotracer is seen in long bones . Cranial vault involvement has been reported in up to 26% of cases .
First-line treatment is considered to be interferon alfa for disease stabilization but given that this medication is not available in Colombia we were forced to consider second-line treatments. Amongst those are anakinra, vemurafenib, imatinib and cladarbine. Our treatment was directed by genetic studies that revealed positive for BRAF V600E proto-oncogene that suggest clonal proliferation giving the possibility of treatment with Vemurafinib and therefore inhibiting the activation of the RAS-extracellular-signal-regulated kinase limiting proliferation and lesion survival . This mutation is present in 55-65% of ECD cases . So far, the patient responded adequately to the treatment and no salvage therapies like vincristine, vinblastine, cyclophosphamide, and methotrexate have been used. Nowadays patients have a 1-year survival rate of 96% and 5-year survival rate of 68% .