INTRODUCTION
Lithium first saw clinical use as a means to reduce uric acid concentrations in patients with gout. The mood-altering effects of lithium led internist Alfred Baring Garrod to erroneously infer the presence of “brain gout” in 1859. In 1871, William Hammond first described the use of Lithium for mania, but widespread adoption did not follow until the mid-20th century.[1]
Despite having decades of proof as an effective treatment for bipolar disorders, lithium use is hampered by a narrow therapeutic index, and toxicity is common. The manifestations of toxicity vary based on acuity of exposure, with gastrointestinal manifestations predominating in acute overdose and neurologic manifestation in chronic accumulation, either by over medication or reduced elimination. Following lithium toxicity, some patients develop prolonged or permanent neurological sequelae, known as the Syndrome of Irreversible Lithium-Effectuated Neurotoxicity (SILENT), which spans a spectrum of neurological symptoms from mild tremors to debilitating cognitive and motor impairments.[2] While the risks for the development of long-term sequelae are not well characterized, there is an association between the duration of elevated serum lithium concentrations (reflecting a high CNS lithium concentration) and an increased likelihood of serious or prolonged symptoms.[3] It follows that expeditious removal of lithium from the body may limit long-term toxicity.
Lithium is well suited for removal by hemodialysis. It is a small molecule, with minimal protein binding and a volume of distribution of 0.7-0.9 L/kg.[4] While it is undisputed that hemodialysis removes lithium, the clinical efficacy of hemodialysis has not been definitively established by well-designed randomized controlled trials. As such, identifying patients likely to benefit from hemodialysis is challenging. The Extracorporeal Treatments in Poisoning Workgroup published a systematic review of the literature in 2015 and established expert consensus guidelines on the use of hemodialysis in lithium poisoning. Despite expert consensus, all suggestions and recommendations were based on very low levels of evidence.[5] The EXTRIP guidelines have never been externally validated.
Seeking to hone these guidelines with the aim of reducing potentially unnecessary treatments, a retrospective analysis of 128 ICU patients with lithium toxicity identified a lithium concentration ≥5.2 mmol/L or creatinine ≥200 µmol/L as indicators for hemodialysis.[6] On a subsequent analysis, these “Paris” criteria were applied retrospectively to the same cohort, and compared to EXTRIP criteria. For each set of criteria, neurological status at ICU discharge was compared between patients who met the criteria and underwent ECTR and those who met the criteria but did not undergo ECTR.[7] The authors conclude that the application of Paris criteria led to a statistical difference in outcome in the primary analysis, and that application of the EXTRIP criteria leads to ECTR in more patients than is necessary, although they acknowledge modifications to the published EXTRIP criteria used in their analysis and the inherent circular nature of validation of the Paris criteria using the derivation cohort.
To our knowledge, neither EXTRIP nor Paris criteria have been independently validated or compared. In this study we applied both EXTRIP and Paris criteria retrospectively to a cohort of patients in a poison center database to determine if differences in neurologic outcomes were significant in cases in which criteria were discordant.