DIAGNOSIS TUMEFACTIVE MULTIPLE SCLEROSIS (MS).
Differential diagnosis based on clinical presentation and neuroimaging included a demyelinating disease, a diffuse intrinsic brain tumor as well as a lymphoma, the latter being rather unlikely due to several normal CSF evaluations (no malignant cells, normal protein). Apart from the fulminant clinical deterioration, atypical features for a demyelinating disease in our patient included the absence of intrathecal immunoglobulin production, the absence of periventricular lesions as well as the mass effect seen on MRI. Reduction of N-acetylaspartate and increase of choline suggesting glial proliferation along with hypermetabolism of the lesions seen on 18-FDG PET left the differential diagnosis of a brain tumor open and urged the histopathological confirmation.
Stereotactic biopsy was challenging in our patient. Notably, the described histopathology resulted from the third sampling after two biopsy attempts in vain, reflecting the technical difficulties concerning stereotactic needle biopsies, especially in the posterior fossa or brainstem (1). Histomorphology definitely excluded CNS lymphoma. The prominent astrocytosis with marked pleomorphism and even atypical mitotic figures argued in favor of a malignant variant of an astrocytoma. Abundant Rosenthal fibers, typically found in pilocytic astrocytomas and Alexander's disease, have been described in both brain neoplasms and in chronic MS lesions of patients with long-standing disease and did not aid in the differential diagnosis (8). The lymphocytic infiltrate found in the stereotactic specimen, which by its nature represented only a small part of the lesion, was too mild for a prototypic early MS lesion. In summary, histopathological investigations could by no means rule out a malignant astrocytoma, which led to the therapeutic scenario mentioned above.
Post mortem pathology finally confirmed a chronic destructive demyelinating disease, compatible with MS. Multifocal lesions were found within the white matter of the brainstem, the cerebellar peduncles and the posterior horn of the left ventricle with central amorphic necrosis surrounded by areas of significant demyelination, astrocytosis and inflammation with microglia cells and macrophages. The necrosis was interpreted as the resulting from the radiation treatment. In cases of atypical clinical symptoms, normal CSF evaluation and uncommon MRI findings biopsies are required to establish the diagnosis of a demyelinating disease (4). Histological characteristics of chronic MS lesions are well known, however, diagnosing an acute, inflammatory demyelinating process in small biopsy specimens (especially stereotactic needle biopsies) can be challenging. The key features of early MS lesions are still demyelination, extensive macrophage invasion, perivascular and parenchymal T cell infiltrates as well as relative axonal preservation. Since early MS lesions are characterized by hypercellularity as well as extensive and prominent astrocytosis they can be confused with astrocytic tumors, especially in small biopsy specimens (2). Reactive astrocytes in MS lesions, that can display significant pleomorphism and even atypical mitotic figures (so-called Creutzfeldt-Peters cells) can be misleading (5). Additionally, the rare coincidence of MS and glial tumors can add to the difficulties of arriving at the correct diagnosis (6). Sending ambiguous biopsy specimens to specialized centers can help in lesion classification (7). Based on the extent of myelin protein loss, oligodendrocyte preservation as well as the composition of the inflammatory infiltrates, different patterns of demyelination have been described, reflecting in part the broad heterogeneity within the clinical presentation of MS (3). Further histological subtyping and characterization of individual pathogenetic patterns may therefore even help in focusing treatment strategies (2). Finally, the insistence on a representative stereotactic specimen is the conditio sine qua non for a clear histological diagnosis, but the presence of abnormal tissue does not necessarily mean the specimen will be diagnostic.
Contributed by Rainer Ehling, MD, William Sterlacci, MD, Hans Maier, MD, Thomas Berger, MD