key: cord-0941487-mdluoeyu
authors: Chung, Ha‐Yeun; Neu, Charles; Wickel, Jonathan; Kuckertz, Sarah L.; Coldewey, Sina M.
title: Neurofilament Light Chain in Patients with COVID‐19 and Bacterial Pneumonia
date: 2021-06-09
journal: Ann Neurol
DOI: 10.1002/ana.26135
sha: f019013a02f7184f7aa24ba64f8939af1165f019
doc_id: 941487
cord_uid: mdluoeyu

nan

NfL value is a well-established marker for neuronal injury. Corroborating our results, recent reports have shown low to intermediate NfL levels in COVID-19 patients as compared to other infectious diseases, for example, bacterial pneumonia and sepsis. 2, 3 Thus, with respect to changes of NfL levels, our own data and current evidence do not indicate commonly occurring neuronal damage in COVID-19. However, differences in cohort composition, such as incidence of delirium or acute kidney injury, could explain the observations made by Sutter and colleagues. As recently reported, delirium itself is associated with NfL elevation and cognitive impairment independent of infection. 4 Furthermore, renal dysfunction might also have influenced NfL levels. 5 In conclusion, we agree with the authors' statement that prospective studies testing the cognitive outcome of COVID-19 patients are needed to evaluate the prognostic value of NfL levels for neuronal injury during acute SARS-CoV-2 infection. Nonetheless, at this stage, we caution against interpreting the NfL data shown by Sutter The authors compared plasma neurofilament light chain (pNfL) levels in patients with bacterial and coronavirus disease 2019 (COVID-19) pneumonia; pNfL levels in bacterial pneumonia 3 days after onset of sepsis were considerably higher than those in COVID-19 pneumonia at days 3 and 7.

Chung and colleagues concluded that their results are corroborated by those of others 3, 4 showing "low to intermediate NfL levels in COVID-19 patients as compared to other infectious dis-eases…" and that their "own data and current evidence do not indicate commonly occurring neuronal damage in COVID-19." Noteworthy, the clinical severity of patients with COVID-19 pneumonia by Chung HY et al were lower with a mean Sequential Organ Failure Assessment (SOFA) score of 4, whereas in our cohort it was 7; 86% of the critically ill patients in our study were ventilated and 17% died (no data are provided about eg, ventilation, comorbidities, intensive care unit [ICU] admission, oxygenation indexes, and outcome in patients with COVID-19 in the cohort of Chung HY et al). In contrast to our cohort, which was analyzed after disease progression that led to an admission to the ICU (ie, representing a later stage of the disease), the patients presented by Chung HY et al were analyzed within the first few days after onset of pneumonia. In fact, the patients with COVID-19 described by us match better with the subgroup categorized as "severe" by Kanberg et al 3 where the median pNfL level was 32.7 pg/ml, very similar to our finding of 36.1 pg/ml in serum (vs approximately 5-10 pg/ml in Fig 1C of Chung HY et al). We consider these levels not as "low to intermediate" as they are in the range of patients with bacterial pneumonia presented by Chung HY et al.

Patients with sepsis-associated encephalopathy show radiological signs of brain damage and neuropsychological signs of brain dysfunction, 4 pNfL levels were strongly increased compared to patients without brain dysfunction; further, they correlated with a poorer long-term neurofunctional outcome. Neuronal damage can be assumed as the cause for the elevation of serum neurofilament light chain (sNfL) observed by Chung et al in their cohort with bacterial pneumonia. Important in our view is the observation that neuronal damage occurs in the course of both COVID-19 and sepsisassociated encephalopathy (ie, also in absence of overt infection of the central nervous system). 4,5 We agree that neuronal damage is not specific for COVID-19, but seems likely a generic consequence in severe infectious disease of various etiologies. We agree as well that the role of renal dysfunction and other metabolic changes as factors modulating NfL levels during infectious diseases needs to be explored. The main focus of the authors was the 34 subjects in Group B with very low dystrophin levels (>0% but <5%). Of these 34, 28 (82%) showed splice site or pseudoexon mutations; these types of mutations are expected to result in residual levels of biochemically normal dystrophin protein.

The authors note that "Very low residual dystrophin protein quantity can cause a shift in disease phenotype from DMD toward BMD" (Abstract). The authors note that their data has implications for therapeutic approaches to dystrophin replacement, such as gene therapy, CRISPR gene editing, and exon skipping.

Unfortunately, there is a key limitation to this interpretation of their data that is not noted by the authors. Namely, the large majority of subjects studied (82%) were likely producing low levels of biochemically normal dystrophin (full-length, 427kDa) from birth. In contrast, there are no current or envisioned therapeutic approaches to DMD that seek to introduce biochemically normal dystrophin. Instead, gene therapy, exon skipping, and envisioned CRISPR approaches aim to introduce biochemically abnormal, semifunctional dystrophin. Although very low levels of biochemically normal dystrophin, especially when present from birth, may mitigate clinical symptoms, this cannot be assumed for biochemically abnormal dystrophin introduced later in life. The study results, although interesting for a genotype-phenotype correlation, should not be extrapolated to being informative in a dystrophin-restoring therapeutic context.

By not making the distinction of biochemically normal versus biochemically abnormal dystrophin, the authors may inadvertently heighten the expectations of patients, families, physicians, and regulatory agencies regarding anticipated clinical benefit from very low levels of semifunctional dystrophin. 1 

Serum neurofilament light chain levels in the intensive care unit: comparison between severely ill patients with and without coronavirus disease 2019

Neurochemical evidence of astrocytic and neuronal injury commonly found in COVID-19

Neurofilament light chain in patients with COVID-19 and bacterial pneumonia

Serum neurofilament light chain levels in the intensive care unit: comparison between severely ill patients with and without coronavirus disease 2019

Neurochemical evidence of astrocytic and neuronal injury commonly found in COVID-19

Very low residual dystrophin quantity is associated with milder dystrophinopathy

Plasma from a subgroup of an ongoing single-center prospective cohort study was employed for analyses (NCT03620409 We thank the other members of the research group Translational Septomics for their support.