A retrospective test for a possible relationship between linezolid‐induced thrombocytopenia and hyponatraemia

What is known and objective Thrombocytopenia is one of the typical adverse events caused by linezolid (LZD). Recently, some cases of severe hyponatraemia occurring while receiving LZD have been reported. This study investigated a possible relationship between LZD‐induced hyponatraemia and thrombocytopenia and identified the risk factors for hyponatraemia and/or thrombocytopenia. Methods In this retrospective, single‐centre, observational cohort study, 63 hospitalized patients aged over 18 years who received intravenous injection of LZD for more than seven consecutive days in Oita University Hospital between April 2015 and March 2018 were analysed. Results Thrombocytopenia occurred in 25 (39.7%) patients and hyponatraemia in 11 (17.5%) patients. Seven of 11 patients with hyponatraemia had concurrent thrombocytopenia. Although both serum sodium level and platelet count declined in most patients who developed hyponatraemia, no significant association between thrombocytopenia and hyponatraemia was found. Creatinine clearance level (Ccr) was significantly lower not only in the thrombocytopenia (vs no‐thrombocytopenia) but also in the hyponatraemia group (vs no‐hyponatraemia group). Univariate and multivariate logistic regression analyses identified different risk factors for thrombocytopenia and/or hyponatraemia (thrombocytopenia: Ccr and administration period; hyponatraemia: serum albumin; thrombocytopenia and hyponatraemia: administration period and serum albumin). What is new and conclusion In conclusion, this study found no significant relationship between LZD‐induced thrombocytopenia and hyponatraemia and identified some possible risk factors associated with onset of the two adverse events. These require further validation. In conclusion, there is no significant relationship between the two adverse events: thrombocytopenia and hyponatraemia, and the changes after discontinuation of LZD as well as the risk factors associated with the two adverse events are possible to be different.