'Radiologically compatible CLIPPERS' may conceal a number of pathologies

Sir, Chronic lymphocytic inflammation with pontine perivascular enhancement responsive to steroids (CLIPPERS) is a newly defined clinical and radiological inflammatory entity prominently involving the pons. Patients with this condition present with an evolving brainstem syndrome and have characteristic punctate and curvilinear gadolinium enhancing lesions ‘peppering the pons’, extending variably into adjacent CNS structures. This condition is said to be responsive to corticosteroids, with patients requiring prolonged steroid or immunosuppressive therapy (Pittock et al., 2010). Here, we report two patients with brainstem syndromes and MRI scans compatible with CLIPPERS. One had a steroid-dependent syndrome, as described by Pittock et al. (2010). However, another—with a typical presentation and MRI scan—had biopsy findings favouring a low-grade glioma and failed to respond to dexamethasone and radiotherapy. This raises the possibility that ‘radiologically compatible CLIPPERS’ may conceal a number of pathologies. A 70-year-old female presented in September 2010 with 1 month history of intermittent rotational vertigo and nausea, fluctuating diplopia and mild gait ataxia. She had a past history of a superficial melanoma excised from her nose in 2006 but was otherwise fit and well. MRI at presentation showed multiple curvilinear enhancing lesions within the pons, with a small number of lesions scattered throughout both cerebellar hemispheres (Fig. 1A). None of the lesions had significant mass effect. Serological investigations were normal. Spinal fluid examination demonstrated a protein level marginally above the upper limit of normal, 0.49 g/l (reference range 0.15–0.45 g/l), otherwise it was unremarkable; in particular, it was negative for oligoclonal bands and no malignant cells were detected. CT scans of her abdomen and pelvis were normal. CT scan of her chest showed an enlarged left lobe of the thyroid gland extending into the superior mediastinum. This was found to be a colloid nodule on ultrasound-guided biopsy. Over the first 6 months, the patient remained symptomatic but stable, apart from the late appearance of hiccups. Her neurological examination revealed only mild gait ataxia. Given the similarity of her imaging and clinical presentation to the patients described by Pittock et al. (2010), she was given a trial of high-dose corticosteroids: 1 g oral methylprednisolone for 3 days, followed by a reducing course of prednisolone—60 mg for 5 days, 40 mg for 5 days and 20 mg for 5 days before stopping. Within a few days of starting treatment, her symptoms improved. However, she experienced a significant relapse following their withdrawal, developing a right intranuclear opthalmoplegia, a right gaze palsy, down beating nystagmus and truncal ataxia. She was re-started on high-dose corticosteroids (500 mg methylprednisolone for 5 days, followed by 60 mg prednisolone reducing over 6 weeks to 30 mg) and again made a rapid response. Her clinical improvement was paralleled radiologically (Fig. 1B). (A) Axial gadolinium-enhanced T1-weighted image at the level of the pons shows multiple small foci of contrast enhancement in the basis pontis, pontine tegmentum and nodulus of the cerebellar vermis. (B) Imaging 5 months later, after corticosteroid treatment, shows substantial reduction in the number of enhancing lesions. A 46-year-old male presented in 2001 with a slowly evolving brainstem syndrome. His past medical history comprised Type II diabetes, hypertension and hepatic steatosis associated with central obesity. Ten months prior to presentation, he developed left facial numbness, followed by painless, horizontal binocular diplopia and mild gait and limb ataxia. Brain MRI (Fig. 2A) showed gadolinium enhancing lesions, predominantly affecting the pons, extending into both cerebral peduncles. Serological investigations were normal. CSF had mildly raised protein (0.69, reference range 0.15–0.45 g/l) but was otherwise unremarkable; in particular, it contained no cells and was negative for oligoclonal bands. CT scans of his chest, abdomen and pelvis were normal. While on no therapy, a pontine biopsy was performed. This demonstrated a lesion comprising atypical enlarged astrocytes having abundant eosinophilic cytoplasm scattered at low-packing density on a fibrillary background featuring Rosenthal fibres (Fig. 3A and B). Numerous cytologically bland lymphocytes, mainly T cells, concentrated around otherwise normal blood vessels were a striking feature (Fig. 3C–E). There was no evidence of vasculitis, lymphoma, sarcoidosis or demyelination. The biopsy findings were described as ‘unusual’, but given the presence of atypical astrocytes and Rosenthal fibres, a low-grade glioma, lacking features of ganglioglioma or pilocytic astrocytoma, was strongly favoured. As a result, the patient was treated with radiotherapy; receiving 54 Gy in 30 fractions over a 6-week period. Five days prior to starting radiotherapy, he was commenced on dexamethasone (2 mg/day); he remained on between 0.5 mg and 8 mg of dexamethasone per day over the next 17 months. His disease did not respond to radiotherapy; he did not improve on higher doses of dexamethasone, nor did his disease worsen with dose reduction. Instead he slowly but relentlessly progressed, clinically and radiologically (Fig. 2B). He died of respiratory failure and aspiration pneumonia in November 2003, 2 years following his initial presentation. No autopsy was conducted. (A) Axial gadolinium-enhanced T1-weighted image at the level of the pons shows multiple enhancing lesions. (B) Imaging 18 months later, after radiotherapy and corticosteroid therapy, shows atrophy of the pons and new enhancing lesions. Neuropathology from Patient 2 showing (A) atypical astrocytes (haematoxylin and eosin stain ×40; scale = 50 μm); (B) Rosenthal fibres arrowed (haematoxylin and eosin stain ×40; scale = 50 μm); (C) parenchyma and perivascular lymphocytic infiltration, without vascular mural infiltration or destruction (haematoxylin and eosin stain ×10; scale = 200 μm); (D) CD3 positive T cells (×40; scale = 50 μm); and (E) CD79a positive B cells (×40; scale = 50 μm). We have reported two cases with clinical and radiological presentations compatible with ‘CLIPPERS’. Our most recent case is indistinguishable from those described in the original Pittock et al. (2010) paper, and is likely to have CLIPPERS, as defined by them. Despite almost identical presenting features, our earlier patient’s diagnosis is most likely to be glioma. Although not definitively diagnostic, his biopsy demonstrated atypical astrocytes and Rosenthal fibres, which were not described in the original CLIPPERS cases. In this case, the key diagnostic feature was that he did not respond to steroids [albeit at lower doses than described by Pittock et al. (2010)], nor to radiotherapy which, perhaps, one would expect for a condition driven by lymphocytes. We suggest, therefore, that radiologically compatible CLIPPERS may conceal a number of pathologies and that steroid responsiveness is a key diagnostic feature of CLIPPERS. A.J.C. and J.L.J. are funded by the Cambridge Centre for Biomedical research.