
Liquid biopsy of cerebrospinal fluid for MYD88 L265P mutation is useful for diagnosis of central nervous system lymphoma
Author(s) -
Yamagishi Yuki,
Sasaki Nobuyoshi,
Nakano Yoshiko,
Matushita Yuko,
Omura Takaki,
Shimizu Saki,
Saito Kuniaki,
Kobayashi Keiichi,
Narita Yoshitaka,
Kondo Akihide,
Shiokawa Yoshiaki,
Nagane Motoo,
Ichimura Koichi
Publication year - 2021
Publication title -
cancer science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.035
H-Index - 141
eISSN - 1349-7006
pISSN - 1347-9032
DOI - 10.1111/cas.15133
Subject(s) - lymphoma , liquid biopsy , primary central nervous system lymphoma , digital polymerase chain reaction , cerebrospinal fluid , biopsy , mutation , medicine , cell free fetal dna , pathology , central nervous system , cancer research , biology , polymerase chain reaction , cancer , gene , genetics , pregnancy , fetus , prenatal diagnosis
The current standard of diagnosing central nervous system (CNS) lymphoma is stereotactic biopsy, however the procedure has a risk of surgical complication. Liquid biopsy of the CSF is a less invasive, non‐surgical method that can be used for diagnosing CNS lymphoma. In this study, we established a clinically applicable protocol for determining mutations in MYD88 in the CSF of patients with CNS lymphoma. CSF was collected prior to the start of chemotherapy from 42 patients with CNS lymphoma and matched tumor specimens. Mutations in MYD88 in 33 tumor samples were identified using pyrosequencing. Using 10 ng each of cellular DNA and cell‐free DNA (cfDNA) extracted from the CSF, the MYD88 L265P mutation was detected using digital PCR. The conditions to judge mutation were rigorously determined. The median Target/Total value of cases with MYD88 mutations in the tumors was 5.1% in cellular DNA and 22.0% in cfDNA. The criteria to judge mutation were then determined, with a Target/Total value of 0.25% as the cutoff. When MYD88 mutations were determined based on these criteria, the sensitivity and specificity were 92.2% and 100%, respectively, with cellular DNA; and the sensitivity and specificity were 100% with cfDNA. Therefore, the DNA yield, mutated allele fraction, and accuracy were significantly higher in cfDNA compared with that in cellular DNA. Taken together, this study highlights the importance of detecting the MYD88 L265P mutation in cfDNA of the CSF for diagnosing CNS lymphoma using digital PCR, a highly accurate and clinically applicable method.