Neurofilament Light and Cognition after Cardiac Surgery: Comment.

We read with interest the recent contribution to your journal from Brown et al., who report an association between a postoperative increase in neurofilament light and worse cognitive outcomes at 1 yr after cardiac surgery.1 These results build on previous work which has shown that a postoperative increase in neurofilament light is similarly associated with an increased risk of delirium,2 and we congratulate the authors on their publication.These findings are of particular importance, as they invite future interventional trials to use change in neurofilament light as a surrogate endpoint for long-term cognitive outcome after cardiac surgery. However, while the authors have adjusted their modeling for various factors and performed several post hoc sensitivity analyses following suggestions from the reviewers, we also wish to draw attention to another potential confounder that may have affected the authors’ results.The relationship between predictive biomarkers of cognitive dysfunction and chronic kidney disease is well recognized: large population-based studies have shown that plasma phosphorylated tau increases in chronic kidney disease, to the extent that individuals with chronic kidney disease require higher cutoff points for predicting changes on cerebral imaging in Alzheimer’s disease.3 While the evidence concerning neurofilament light is limited to smaller studies, a similar, positive correlation between blood levels of neurofilament light and creatinine—even after adjustment for age, sex, and body mass index—was observed.4 It is therefore possible that altered renal function may explain changes in neurofilament light, independent of any postoperative brain injury.Acute kidney injury (AKI) is common after cardiac surgery, and the development of AKI has been associated with postoperative delirium in cardiac surgical patients.5 The authors have excluded patients with end-stage renal failure, and the adjustment of their modeling for EuroSCORE-II, cardiopulmonary bypass time, and anemia, which will have mitigated the effect of postoperative AKI on measured neurofilament light to some extent.6,7 However, a further statistical correction for postoperative day 1 serum creatinine or estimated glomerular filtration rate could be used to determine whether change in neurofilament light and its association with worse longer-term cognitive outcomes are independent of postoperative AKI, and reinforce its validity as an endpoint in future clinical trials.While the authors may be justifiably concerned about the risk of multiplicity, we hope that they will consider performing such a post hoc analysis and informing us of their findings.Funding for Drs. Miles and Sanders was received solely from institutional and departmental sources. Dr. Lankadeva is supported by a Future Leader Fellowship of the National Heart Foundation of Australia (Canberra, Australia). Dr. Ayton is supported by an L1 Investigator Grant from the National Health and Medical Research Council of Australia (Canberra, Australia).The authors declare no competing interests.