Convulsions in a critically ill patient on hemodialysis: Possible role of low dose colistin.

The increasing prevalence of multi-drug resistant Gram-negative pathogens in intensive care units has led to the revival of colistin. Colistin had gone into disrepute in early 1970s because of numerous reports of adverse renal and neurological effects. The renewed interest in colistin has also revived the discussion about its toxicity. The neurotoxicity reported in literature is usually with higher doses of colistin. We present a case report of seizures in a critically ill-patient, possibly with low dose colistin. A 47-year-old hypertensive female with chronic kidney disease-5 with sepsis on colistimethate sodium 1 million units (80 mg), intravenous once daily, developed paresthesias and seizures on 12(th) day of therapy, which were subsequently controlled after withdrawl of the drug. To conclude, colistin should be considered as a cause of convulsions in critically ill-patients with renal failure, even when given in low dose and patient receiving intermittent hemodialysis, when other obvious causes have been ruled out. When possible, cessation of therapy may be considered.

Introduction

The increasing prevalence of multi-drug resistant (MDR) gram-negative pathogens in intensive care units (ICUs) and shortage of new antibiotics to combat them has led to the re-evaluation of colistin.[1] Colistin is a multicomponent polypeptide antibiotic, comprised of colistins A and B, which became available for clinical use in the 1960s.[2] It had gone into disrepute because of numerous reports of adverse renal and neurological effects.[13] The renewed interest into colistin has also revived the discussion about its toxicity. We present a case report of possible neurotoxicity with low dose colistin presenting as seizures in a critically ill patient.

Case Report

A 47-year-old known hypertensive female with chronic kidney disease-5 [CKD-5], on thrice weekly hemodialysis (HD), presented with cough with expectoration and breathlessness for 3-4 weeks and fever for 10 days. She was managed in an outside facility for 1 week on amlodipine, erythropoietin, imipenem, HD and other supportive care for CKD. On presentation, patient was conscious, well-oriented with a heart rate of 116 beats/min regular, non-invasive blood pressure of 108/66 mmHg, temperature of 100.2°F and respiratory rate of 34/min with accessory muscles being used. On chest examination, bilateral normal vesicular breath sounds were heard with decreased air entry and occasional coarse crepts in the right lower zone. Her laboratory workup on the day of admission and subsequently is shown in Table 1. Chest X-ray showed right lower lobe consolidation/collapse with pleural effusion. Blood and urine cultures were sent, which were sterile. She was managed in the medical intensive care unit on antibiotics and non-invasive ventilation. She was continued on injection imipenem 500 mg twice daily (BID), which she had already received for 2 days prior to admission. On the 3rd day of admission, the patient's clinical condition deteriorated with a significant rise in total leukocyte count (TLC) up to 28,600/mm3. Repeat blood cultures were sent and injection teicoplanin and injection caspofungin added empirically. Next day, a provisional report of gram-negative coccobacilli in blood was received, which was later confirmed to be Acinetobacter baumannii. Injection colistin in a dose of 1 million units (MU) (80 mg colistimethate sodium) intravenous (i/v) once daily (OD) following a loading dose of 2 MU was started. Her fever and TLC started decreasing gradually from the 7th day onward. Caspofungin was de-escalated on receiving culture reports and imipenem was stopped after 10 days of therapy. Patient continued to improve clinically with off and on low-grade fever. She was continued on almost alternate day HD.

Table 1

Serial laboratory investigations

On the morning of 16th day of admission, patient had sudden onset of abnormal facial twitchings, which were mainly circumoral and initially limited to the neck. The seizures were controlled with short acting benzodiazepine in the form of injection midazolam 1 mg i/v stat. She was loaded with injection phenytoin 1 g i/v on neurologist's opinion, followed by 100 mg i/v BID. But in the same evening, patient had another episode of seizures, which were generalized tonic clonic and got relieved with injection midazolam 1 mg + 1 mg i/v bolus. Levirecetam 1 g i/v stat followed by 500 mg i/v BID was added by the neurophysician. Her laboratory workup reflected no acute metabolic derangements. Liver profile was normal. Neuroimaging of the brain (magnetic resonance imaging) and electroencephalography were normal. Nerve conduction velocity to rule out critical illness neuropathy was also done. To rule out meningitis as a cause for fever and seizures, lumbar puncture was performed and the cerebral spinal fluid (CSF) was found to be clear, with normal pressure and CSF biochemistry was within normal limits (CSF glucose = 56 mg/dl (corresponding Random Blood Sugar = 92), proteins = 37, TLC = nil).

Further search for any offending drug as a cause for seizures was made. Imipenem which is a potentially seizure inducing drug, was stopped around 1 week before the occurrence of focal seizures. Excluding all other possible causes, colistimethate which is a seizure-threshold lowering drug, was considered to be the sole possible causative agent for seizures. As the patient had improved clinically, colistin was stopped on the occurrence of seizures. Fresh blood cultures were sent after 48 h of stopping antibiotics, which were sterile. Patient had no recurrence of seizures at any point of time after stopping the drug. The antiepileptics were tapered after 1 week of starting the treatment. The patient was shifted to the ward and was finally discharged in a fair condition after 33 days of hospital stay.

Discussion

After being abandoned in the 1970s because of reported nephrotoxicity and neurotoxicity, colistin is having a second life as a “salvage” treatment against multiresistant gram-negative bacteria (GNB).[14] Colistin belongs to the polymyxin class of cationic polypeptide antibiotics, composed of at least 30 different polymyxin compounds, mainly colistins A and B.[5] The mounting prevalence of MDR GNB worldwide has renewed interest into colistin, but it also has revived the discussion about its toxicity.[1] The incidence of colistin-associated nephrotoxicity and neurotoxicity in recent literature is low.[3] The reported adverse effects in older studies were usually observed with high doses of colistin.[456] Paradoxically, the present case report considers even low-dose colistin as a cause for focal and generalized tonic clonic seizures. In a known case of CKD with sepsis, multiple factors can lead on to seizures including metabolic derangements, dialysis disequilibrium syndrome, sepsis itself and concomitant drug therapy. In our patient, we had ruled out all possible causes of seizures on laboratory work-up and radiological studies. A potentially seizure inducing drug imipenem had been stopped 8 days prior to the episode of seizures. The plasma half-life of imipenem/cilastatin is approximately 1 h in a patient with a normal renal function, which may increase to more than 3 h with imipenem and 8.84-17 h with cilastatin in deranged renal function.[78] But both imipenem and cilastatin are cleared from the circulation by HD and half-lives are shortened.[8] Thus, it is highly unlikely to remain in circulation after three HD sessions which our patient had undergone after stopping imipenem. Hence seizures in our patient could not be attributed to imipenem. Dialysis disequilibrium can be another possibility in our patient, but usually it is more often seen when patients are undergoing their first HD. In our case, patient was already dialysis-dependent.

No clear-cut specifications on colistin dosage are available and colistin has been used in different doses in patients with normal or deranged renal function.[23] Colistin pharmacokinetics (PK) are expected to be dramatically altered in critically ill patients, because they are prone to large swings in distribution volume, fluctuations in renal clearance and variable protein binding.[259] Very limited information is available on the PKs of colistimethate and colistin in critically ill patients receiving renal replacement therapy. In renal failure, renal excretion of colistin methanesulfonate (CMS) is decreased resulting in higher conversion to colistin and prolongation of half-life.[9] But colistin is mainly excreted by non-renal mechanisms.[9] In a patient undergoing continuous venovenous hemodiafiltration (CVVH), conversion of CMS to colistin was rapid and terminal half-lives of CMS and colistin were 6.8 h and 7.5 h, respectively.[10] As our patient was also on alternate day HD and PK of colistin is unpredicted, the causal role of low-dose colistin with seizures in our patient could not be established with certainty. The removal of CMS and colistin by dialysis has yielded conflicting results in the literature.[29] Li et al. retrieved colistin and its CMS derivative from plasma and dialysis fluid during CVVH using high-performance liquid chromatography, suggesting that dose adjustment is smaller in these patients.[11] Sarria et al. failed to detect polymyxin B in the dialysate during CVVH; although, its presence in the plasma had been confirmed, suggesting that the drug is not dialyzable and therefore supplemental doses would not be necessary in patients subjected to dialysis.[12] However, it is quite likely that the daily dose of colistin might exert an important influence on adverse effects and numerous studies report that neurotoxicity increases strikingly with the mean daily dose administered and total cumulative dose.[134613] In one of the studies by Koch-Weser, neurological reactions rate reached 45% among patients who received over 400 mg/day of colistin.[3] Sabuda showed varying neurological signs and symptoms with a dose of colistin varying from 150 mg every 36 h to 125 mg every 6 h.[14] Another case report by Wahby described neurotoxicity with 170 mg BD colistin base within 5 days of initiation of treatment.[15] However, our patient was on a low dose colistimethate sodium 1 MU (80 mg) i/v OD based on the creatinine clearance (Cr Cl < 10 ml/min).[16] This was a very low dose as compared to other studies quoting neurotoxicity with colistin, even with deranged renal parameters.[346131415] From our review of literature, we could not find colistin neurotoxicity with dose lesser than this.

Colistin neurotoxicity may be associated with dizziness, muscle weakness, facial and peripheral paresthesia, partial deafness, visual disturbances, vertigo, confusion, hallucinations, seizures, ataxia and neuromuscular blockade, paresthesias being the most frequent.[1315] However, the most dreaded neurotoxic event is neuromuscular blockade presenting as a myasthenia-like syndrome or respiratory muscle paralysis producing apnea.[414] Diagnosis of neurotoxicity is mostly made on clinical grounds, making it difficult to discriminate from “critical illness polymyoneuropathy” in ICU patients.[4] The interaction of colistin with neurons having high lipid content has been postulated as a cause of neurotoxicity.[3] Other investigators have suggested a biphasic mechanism to explain the neurotoxic event; a short phase of competitive blockade between acetylcholine and polymyxins is followed by a prolonged phase of depolarization associated with calcium depletion.[1] Potential triggers of neurotoxicity are hypoxia, concomitant medication (muscle relaxants, narcotics, sedatives, anesthetic drugs and corticosteroids) and impaired renal function.[4] The female gender predisposition in neurotoxicity, has also been seen in our patient.[13]

Mild neurological complications of colistin such as paresthesias usually subside after prompt cessation of the offending drug.[117] We stopped colistin in our patient very next day of seizure episode and thereafter, there was no recurrence of seizures even on tapering off the epileptic drugs. For severe neurotoxicity, use of i/v calcium and choline-esterase inhibitors, e.g., neostigmine and edrophonium, have been recommended with conflicting results and in patients with coexisting renal failure, HD has also been indicated.[18]

Conclusion

Although rare, but the toxicity potential of colistin should be kept in mind while prescribing the drug especially in critically ill-patients in ICUs, in whom colistin PK are expected to be dramatically altered. Colistin being a cause of seizures in this patient is not definitely proven, though highly probable. Colistin should be considered as a cause of convulsions in critically ill-patients with renal failure, even when given in low dose and patient receiving intermittent HD, when other obvious causes have been ruled out. When possible, cessation of therapy may be considered. Although considering the spectrum of colistin usage in ICU during recent times, it might not be possible to stop colistin altogether, Studies are required to determine if monitoring of serum colistin levels in ICU can prevent toxicity.