Background: A 40-year-old man with chronic myelogenous leukemia presented multiple times over a period of 3 years with episodes of confusion, wide-based gait and falls because of recurrent hydrocephalus despite repeated therapeutic lumbar punctures. These problems occurred in the context of persistent cerebrospinal fluid (CSF) pleocytosis and leptomeningeal enhancement. Extensive diagnostic workups and therapeutic trials had failed to identify a clinically plausible cause or produce any significant improvement in the CSF and neuroimaging abnormalities. Methods: We used high-throughput metagenomic shotgun sequencing to identify microbes in 2 CSF samples collected from the patient during his illness. These results were compared to sequence data from 1 CSF sample collected during treatment and 5 control CSF samples from other patients. Results: We found sequences representing 53% and 67% of the Propionibacterium acnes genome in 2 CSF samples collected from the patient during his illness. Directed antimicrobial therapy was administered for 6 weeks with resolution of CSF and neuroimaging abnormalities. Sequencing of a sample obtained during treatment demonstrated that the P. acnes levels were decreased to background levels. After insertion of a ventriculo-peritoneal shunt, the patient returned to baseline status. Conclusions: High-throughput metagenomic shotgun sequencing revealed P. acnes as the cause of chronic meningitis that had eluded conventional attempts at diagnosis. Treatment directed at this organism resulted in cure of the infection and clinical improvement.

Original languageEnglish
Pages (from-to)227-233
Number of pages7
JournalTransplant Infectious Disease
Issue number2
StatePublished - Apr 1 2016


  • High-throughput sequencing
  • Meningitis
  • Metagenomic sequencing
  • Propionibacterium acnes


Dive into the research topics of 'High-throughput sequencing of cerebrospinal fluid for diagnosis of chronic Propionibacterium acnes meningitis in an allogeneic stem cell transplant recipient'. Together they form a unique fingerprint.

Cite this