Our case report highlights some important issues for consideration regarding the use of ertapenem in CKD 5D patients. In our case, the presentation of ertapenem-induced neurotoxicity manifested as visual and auditory hallucinations and agitation. The Naranjo Scale, which estimates the probability of an adverse drug reaction, was used to evaluate this potential causality. Our patient scored 6, which suggested probable causality [7]. Even though we were unable to obtain ertapenem drug levels during admission to assess the appropriateness of his dosing, the resolution of symptoms after discontinuation of ertapenem, and absence of other significant possible causes, indicated that it was high likely that his symptoms were due to ertapenem-induced encephalopathy. In clinical trials of ertapenem, the main neurological adverse events were headaches (2.2%) and seizures (0.2–0.5%) [8].

Hallucinations and acute confusional state, as seen in our case, have been described in other case reports [4, 5].

Ertapenem has high lipophilicity and its half-life is significantly increased in CKD patients especially those with low eGFR. Furthermore, only 30% of ertapenem is filtered during HD. Current dosing advise for dialysis patient were originated from old studies using low flux dialyser. High flux dialyser HD and HDF enhanced the clearance of drug as it able to remove larger molecule efficiently and may contribute to subtherapeutic level in CKD5D patient [9]. However, this might not be the reason in our case. Clearly there are other determinants that affected drug clearance in our patient.

Hypoalbuminemia, which is commonly seen in end-stage renal disease (ESRD), is a contributing factor that leads to increased free drug concentration and hence increases central nervous system (CNS) exposure. In few reviews of ertapenem-induced neurotoxicity in humans, the onset of symptoms varies from 4 to 10 days. However, patients with CKD 5D were found to have earlier onset (< 5 days) [4, 8].

In a case series study, in which ertapenem concentration was measured in one CKD 5D patient, it was found that 12 h post 500 mg IV ertapenem the plasma level at the beginning of HD was much higher than the minimum inhibitory concentration (MIC) 90 level. A further increase was seen before the second session of HD and the patient developed CNS symptoms during the fourth HD session (after seven consecutive doses) [3]. This contrasts with our patient, where the onset was later on day 13 of ertapenem. This difference in the onset of symptoms may be due to lack of awareness of the symptoms, and other contributing factors in our patient such as older age, low albumin and morbid obesity. In a retrospective study of ertapenem-induced seizures among CKD 5D patients, male sex, dementia and concomitant use of antibiotics were found to be the main predictors of seizure [1].

The current approved ertapenem dosing guidelines recommend an adjusted dose for CKD 5D patients. This dosage is based on pharmacokinetic data and has not been clinically studied. In a study on the AUC0-∞ of 1 g daily ertapenem in CKD 5D patients, it was found that the free drug concentration was five folds higher than in patients with mild renal impairment at eGFR > 30 [2]. This could be excessive as identified in our case and several other case reports [1, 10]. A recent pharmacokinetic study of the effects of ertapenem when administered as a 500 mg dose three times weekly found that a plasma concentration above 2 mg/L was maintained (selected based on the MIC90 and the CLSI/FDA susceptibility breakpoint) [11].

This finding is exciting and could be of practical benefit for CKD 5D patients, where a three times weekly dose could be administered via a HD line, thus conserving other vessels for future HD access.

This case draws attention to the risk of potentially serious CNS toxicity in HD patients who receive the current recommended dose of ertapenem. Increased awareness, prompt diagnosis and immediate discontinuation are all crucial to avoid unnecessary investigations and potentially appalling complications. Importantly, this report highlights that renal dosing in CKD 5D patients’ needs to be studied clinically and pharmacokinetically to ensure the antibiotic safety.