Blood and CSF monitoring of cefepime-induced neurotoxicity: nine case reports

Sir, Cefepime is a parenteral fourth-generation cephalosporin commonly used as a first-line empirical treatment for severe infections such as pneumonia or septicaemia, especially in neutropenic patients. Neurotoxicity is a well-known adverse effect of cephalosporins and, for cefepime, has been mainly reported in patients treated with high dosages with renal failure. Cefepime-related neurotoxicity reportedly affects 3% of treated patients, although this may be an underestimate.1 Over the last 2 years, six cases of cefepime neurotoxicity were documented in our hospital by demonstration of excessive drug concentrations. Here, we describe these six cases and discuss three cases we have previously published.2,3 Biological fluid assays of cefepime were performed by HPLC for the nine patients with suspected neurological events. The patients’ characteristics are described in Table 1. In all patients, cefepime overdosage was confirmed by analysis of cefepime blood concentrations. After withdrawal of cefepime, the patients’ neurological status improved in all cases. In case 3, cefepime treatment was continued after dose adjustment. In five cases, a decrease in cefepime concentration was observed during neurological improvement. All patients had chronic kidney disease (CKD) or acute kidney disease. Four patients with CKD underwent supplementary dialysis sessions in order to decrease the cefepime concentrations. CSF was available for two patients; measurement of cefepime concentrations in CSF confirmed brain accumulation. Characteristics of cefepime-induced toxicity in our case series AKD, acute kidney disease; CKD, chronic kidney disease; ESRD, end-stage renal disease; HD, haemodialysis; NA, not available. Characteristics of cefepime-induced toxicity in our case series AKD, acute kidney disease; CKD, chronic kidney disease; ESRD, end-stage renal disease; HD, haemodialysis; NA, not available. The proconvulsive action of cephalosporins may be related to a suppressive effect on inhibitory neurotransmission. Sugimoto et al.4 have suggested that β-lactam antibiotics may induce convulsions via a concentration-dependent inhibition of GABAA-mediated neurotransmission, leading to a decrease in seizure threshold. This hypothesis is supported by the observation that benzodiazepines were effective in treating convulsions in most reported cases of cephalosporin neurotoxicity (including cases 2 and 7 in our series). Other mechanisms involving tumour necrosis factor-α or enhanced glutaminergic activity have also been suggested.1 Cefepime is mostly eliminated unchanged via glomerular filtration, and its renal clearance is closely related to CLCR.5 This justifies dose adjustment according to renal function. However, in patients with renal disease, CSF cefepime concentrations may also be increased in relation to increased blood–brain barrier permeability, with decreased serum protein binding and accumulation of organic acids leading to competitive inhibition of active transport from CSF to blood.1 Thus, when CSF is available, determination of cefepime concentration may be of clinical value. The CSF/blood ratio is usually around 10%.1 In cases 2 and 7 in our series, this ratio was increased to 45% and 34%. In case 2, the latency of the adverse effect was longer than that usually described. It can be hypothesized that renal failure deteriorated slowly, resulting in progressive cefepime accumulation until toxic levels were reached. As with most cases published in the literature, the cases reported here concern patients treated with high drug dosages with renal failure, especially true in cases 2, 5, 7 and 8 where the doses used, in order to treat severe infections, were too high in terms of renal function. However, cefepime-induced neurotoxicity has also been reported after dose adjustment to renal function (according to the Summary of Product Characteristics).2 Lastly, some cases have also been described in patients who developed acute renal failure (in response to concomitant treatments or clinical events) during cefepime treatment (as in cases 2 and 8 of the present series). When overdosage is suspected, cefepime treatment should be discontinued, as, in this series and in most published cases, discontinuation of cefepime, either alone or associated with antiepileptic therapy, led to complete resolution of the neurological events. Due to the low drug–protein binding of cefepime, a 3 h haemodialysis session is efficient to remove 70% of a given dose. Many published cases have reported recovery of normal neurological status after haemodialysis.1 Cefepime overdosage may sometimes be difficult to diagnose, since treated patients often present comorbidities that could at least partially account for the neurological symptoms. Thus, monitoring of renal function and cefepime blood concentrations appears to be of value in patients with renal failure, elderly patients and those with a history of neurological disease. Target trough concentrations of cefepime have not been well established. Lamoth et al.6 have described an association between high trough concentrations and neurological adverse events and predicted that a concentration ≥22 mg/L has a 50% probability of inducing neurotoxicity. The trough concentrations in the present series were all ≥22 mg/L, and clinical improvement was associated with a decrease in blood concentration. Our series confirms that cefepime accumulation can be associated with neurological adverse effects. Most cases occur in patients with renal failure and sometimes despite dose adjustment for renal function. Monitoring of blood concentration therefore appears to be a useful tool to avoid or confirm cefepime overdosage and to monitor neurological improvement. Determination of CSF cefepime concentrations may also help to confirm the role of cefepime in neurological events. This study was carried out as part of routine work at the Department of Clinical Pharmacology at Amiens Picardie University Hospital. None to declare.