CNS nocardiosis has an estimated mortality of 20-55% and delayed diagnosis is common. Traditional empiric antibiotics used for bacterial CNS abscesses do not cover Nocardia species.  

Authors aim to describe the clinical and pathological characteristics of patients with central nervous system (CNS) nocardiosis in an effort to determine when empiric coverage for Nocardia should be added to traditional empiric antibiotics for CNS abscesses. 

In this retrospective case series, patients with CNS nocardiosis treated at the University of California San Francisco from 1998-2017 were identified through a comprehensive institutional radiologic database and reviewed using electronic health records. 

Fourteen CNS nocardiosis cases were identified, of whom 9 were male (64%) and 6 Caucasian (55%). Average age on diagnosis was 61±14 years. The majority of patients were immunosuppressed (N=10, 71%), most commonly from chronic prednisone (N=7, 70%) but also due to malignancy (N=5, 50%), organ transplant (N=4, 40%) and autoimmune disease (N=1, 10%). Common chronic comorbidities included diabetes (N=5, 36%) and kidney disease (N=4, 29%). Nearly all patients had evidence of active pulmonary infection during hospitalization (N=12, 86%). Patients with CNS nocardiosis were diagnosed a median of 7 days from admission (range 2-124 days) by culture, except one specimen that was identified by PCR and next-generation sequencing. Most patients underwent neurosurgical abscess drainage (N=10, 71%). Blood cultures were positive for Nocardia in only one case. 

In this case series of patients with CNS nocardiosis, the majority of patients were immune suppressed, and almost all exhibited concomitant pulmonary infection at the time of diagnosis. Blood cultures were often unhelpful and neurosurgical specimen provided the best source of diagnostic confirmation. Clinicians should consider starting empiric therapy for CNS nocardiosis in patients presenting with concurrent brain abscesses and pulmonary infection, particularly those who are immune suppressed.