• valvular heart disease
• clinical pharmacology
• infectious diseases
• neurology

A 76-year-old heavy smoker with vascular disease, mitral valve repair and creatinine 1.4 mg/dL was admitted with fever. Blood cultures grew Enterococcus faecalis and echocardiography demonstrated mitral and aortic valve vegetations. Ceftriaxone (2 g twice daily) and ampicillin (3 g three times a day) were started with defervescence, but creatinine slowly rose to 3.3 mg/dL. On treatment day 14, he became agitated, confused, then comatose (Glasgow Score 4/15) without localising neurological signs. No abnormality to explain the coma could be found. Cerebrospinal fluid (CSF) was normal. Haemodialysis had no effect. Neuroimaging was unremarkable. Electroencephalography (EEG) demonstrated triphasic waves (TWs) (figure 1). Reviewing his drug treatment, ceftriaxone-associated encephalopathy appeared the only suspect.1 Ceftriaxone was stopped. Within 48 hours, he dramatically improved regaining consciousness.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The patient’s electroencephalography showing triphasic waves (highlighted), more prominent in the frontal areas, with predominant theta and delta background activity.

Our patient’s severe reversible encephalopathy was probably caused by ceftriaxone (Naranjo Scale 5–8). These types of adverse events are rare and the potential of antibiotics to mediate acute encephalopathy (antibiotics-associated encephalopathy, AAE) remains poorly recognised.

A recent comprehensive literature search identified only 391 reported patients2; 18% had cephalosporin-associated encephalopathy, mostly related to cefepime or ceftazidime. Ceftriaxone was involved in <10 patients. Clinical features exhibit a remarkable phenotypic variety, including myoclonus, seizures (occasionally non-convulsive) or psychosis. At opposite ends of this spectrum, mental status changes and coma can be found.1 Onset days after antibiotic initiation, normal neuroimaging, abnormal EEG and quick resolution upon drug discontinuation (median 5 days) were characteristics.2 Importantly, >66% of the patients had renal impairment.3 Other risk factors are excess drug dosage and prior brain disease.3

Our patient gradually lapsed into coma while receiving 4 g/day ceftriaxone when glomerular filtration rate (GFR) dropped to 17 mL/min/1.73 m. Renal impairment could be due to acute cephalosporin-associated interstitial nephritis.3 The association of coma with ceftriaxone is supported by the normal neuroimaging, the EEG demonstrating TWs in the absence of other causes4 and remarkable complete reversibility (<48 hours) after drug withdrawal. Haemodialysis had no effect on the coma, also consistent with AAE, since ceftriaxone is poorly dialysed.

The pathogenesis of AAE is incompletely understood. The main mechanism of neurotoxicity involves gamma-aminobutyric acid (GABA)-A receptor. Β-lactams can bind to the receptor impeding inhibitory neurotransmission. The increased excitatory neuronal activity, also due to cytokine release, is the likely pathophysiological basis of Β-lactam-associated encephalopathy.5 Indeed, cephalosporins with high affinity for GABA-A receptors and those which better penetrate the blood-brain barrier (BBB) were found to be more neurotoxic in animal models.

Thus, with the wide use of ceftriaxone, the possibility of antibiotic-associated encephalopathy should be recognised and AAE considered in patients who develop any unexplained neurological symptoms during ceftriaxone administration, especially if their renal function is significantly impaired.