ML-7 Liquid biopsy for MYD88 mutation in cerebrospinal fluid in patients with suspected primary CNS lymphoma

Abstract

Background: Treatment intervention for central nervous system lymphoma (CNSL) requires pathological diagnosis by surgical biopsy. However, there are some cases in which the risk of surgery is high due to age, comorbidities, localization of lesions, etc. We are developing a CNSL diagnostic method based on the detection of MYD88 L265P mutation by digital PCR (dPCR) using CSF-DNA, and a high accuracy with a sensitivity of 92.9% and a specificity of 100% has been reported. Here, we report two cases with suspected brain stem CNSL, whose treatment strategy was determined by integrated clinico-laboratory information including neurological presentations, imaging, and the result of liquid biopsy. Result: Case 1. A 63-year-old woman visited our hospital with a complaint of right hemiplegia, which deteriorated in two months. MR images revealed a contrast-enhancing lesion in the left midbrain-ventral pons, suggesting CNSL. Biopsy was not considered because of its location, while dPCR using CSF-DNA showed a cluster of MYD88 mutation signals. Based on these work-ups, she was treated with high-dose methotrexate-based chemotherapy, resulting in a complete response with marked improvement of symptoms. Case 2. An 83-year-old man was referred for a history of diplopia and ataxic gait lasting for a month. MR images revealed an invasive lesion on his right midbrain-dorsal pons. Biopsy was declined due to the location, and liquid biopsy using CSF-DNA was performed to assist the diagnosis. In the first test, the CSF-DNA yield was too insufficient to determine the mutation signal by dPCR. The second dPCR using sufficient amount of CSF-DNA resulted in the Target/Total value of 0.049% which was lower than the threshold, suggesting the absence of MYD88 mutation. The patient underwent radiation therapy accordingly.Conclusions: CSF MYD88 mutation analysis by dPCR may have clinical utility and requires sufficient amount of CSF-DNA for exclusion of noise signals.