Sirs: anti-Yo or anti-Purkinje Cell Antibody (PCA) is one of the commonest onco-neural antibodies recognized within the CNS. The associated paraneoplastic cerebellar degeneration (PCD) is a clearly defined syndrome with over 200 reported cases in the literature [6]. Magnetic resonance imaging (MRI) is “remarkably unremarkable” in anti-Yo associated PCD although cerebellar atrophy can be seen early in the course of the disease [7, 8]. The negative imaging in anti-Yo positive PCD contrasts PCD associated with other onconeural antibodies, notably anti-Hu but also with anti-Ma2, anti-CV2, and with antibody negative PCD. In such cases PCD occurs in the setting of a more widespread encephalomyelitis and may demonstrate multifocal MRI and functional imaging abnormalities [2, 3, 5]. To date there are no reports demonstrating MRI evidence of cerebellar or extra-cerebellar inflammation in a living patient with anti-Yo associated PCD. A 62 year-old right-handed woman presented with a two-week history of predominantly left-sided limb ataxia, dysarthria and weight loss. Over four weeks in the hospital, her ataxia became generalized with symmetrical limb involvement, horizontal nystagmus, truncal ataxia, and scanning dysarthria. She had normal cognition and no evidence of sensory neuropathy. Initial MRI brain scan was normal. CSF demonstrated a lymphocytic pleocytosis (WBC 9/mm3) and oligoclonal bands. Despite normal serum tumor markers and clear CT imaging of chest, abdomen, and pelvis, Positron Emission Tomography (PET) revealed a hypermetabolic area in the left pelvis, which was subsequently investigated at laparoscopy. Anti-neuronal antibody screening was positive for anti-Yo antibodies and negative for anti Hu, Ri, Ma, Tr and PCA2. A small poorly differentiated ovarian adenocarcinoma was found and bilateral salpingo-oophorectomy performed. A second MRI scan was performed six weeks after onset of symptoms due to clinical deterioration with anti-neuronal antibody status yet unknown. This demonstrated an abnormality in the right medial temporal lobe. This was hyperintense on FLAIR and enhanced with gadolinium (Fig. 1). No cerebellar abnormalities were seen. Following surgical treatment of the tumor, a follow-up scan (at 12 weeks) showed resolution of the medial temporal lobe changes but demonstrated a focal right cerebellar hyperintensity on gadolinium enhanced T1-MRI (Fig. 1). There was no meaningful clinical improvement in her ataxia following surgery and chemotherapy for the tumor. This patient’s case is clinically typical of anti-Yo associated PCD demonstrating a purely cerebellar syndrome with subacute progression resulting in severe disability over 10 weeks. The evolving MRI abnormalities are atypical for the disorder. We believe they represent inflammation. Inflammatory changes have been demonstrated radiologically in other reports of PCD but never in association with anti-Yo [3, 5]. It has been suggested that an inflammatory phase precedes cerebellar atrophy in PCD and that this may even precede clinical presentation [8]. The MRI of the brain was normal at two weeks from symptom presentation. The follow-up scan at six weeks showed right mesial temporal abnormality but no cerebellar change. MRI following resection of the tumor showed resolution of the temporal abnormality but a new focal cerebellitis. No immunosuppressive drugs were given over this period. Although non-inflammatory etiologies such as infection, ischemia, or metastasis could account for the MRI lesions of the cerebellum and temporal lobe in a patient with cancer, the appearance and evolution of this patient’s MRI abnormalities are not consistent with such differentials. Pathologic change in anti-Yo associated PCD is generally restricted to the cerebellum with cortical atrophy and selective Purkinje cell loss. Inflammatory cells may occasionally be seen in the leptomeninges or in the dentate nucleus. In one post mortem study inflammatory changes were noted in the medulla and pons, corresponding to the known expression of the PCA/Yo antigen (a 62 kDa cytoplasmic protein) in cerebellar and brainstem neurons [4, 9]. Therefore it is possible that anti-Yo could be responsible for mesial temporal inflammation. LETTER TO THE EDITORS
Read full abstract
Jan 1, 2008
Kazumasa Hayasaka + 3
Open Access