LAUSR

We read with great interest the article published by Dalla Costa et al concerning the role of serum neurofilament light chains (NfL) as candidate biomarkers of progressive multifocal leukoencephalopathy (PML) during natalizumab treatment (NTZ-PML) in multiple sclerosis (MS) patients. The authors found a 10-fold increase in serum NfL levels at PML onset (n = 25) compared either to pre-PML samples (n = 10) or controls (natalizumabtreated [n = 136] and untreated [n = 151] MS patients), using an electrochemiluminescence assay. In light of these findings, we retrospectively measured serum NfL levels of 5 NTZ-PML patients at our center using the ultrasensitive SiMoA technology. According to the current PML diagnostic criteria, 1 patient met the criteria for definite PML and 3 patients met the criteria for probable PML. In 1 patient, PML was diagnosed based on the typical lesions and evolution of PML lesions on magnetic resonance imaging (MRI). None of the patients had a biopsy. Remarkably, 4 patients were asymptomatic at time of diagnosis. One patient experienced no clinical symptoms during the entire disease course; 4 patients who suffered from neurological symptoms partially recovered, with current Expanded Disability Status Scale scores ranging from 3.5 to 6. No deaths occurred in our cohort. Serum samples were taken at the start of natalizumab treatment and at regular intervals, enabling us to longitudinally test NfL levels in pre-PML samples. In addition, we assessed whether serum NfL correlated with PML lesion volume on MRI at PML onset. PML lesion volume on MRI involved all lesions attributed to PML; the methods of measuring PML lesion volume have been described elsewhere. In the Figure, A shows that in 4 of 5 patients, NfL levels increased at PML onset. In 2 patients, the increase in NfL commenced prior to PML diagnosis. Comparing the mean pre-PML NfL levels (samples between 5 months after natalizumab initiation and 5 months prior to PML diagnosis) to the NfL levels at time of diagnosis of PML, there was an 11.6-fold increase in serum NfL in Case 1, a 3.1-fold increase in Case 2, a 2.0-fold increase in Case 3, a 0.9-fold decrease in Case 4, and a 2.5-fold increase in Case 5. After correcting for age, serum NfL correlated exceptionally strongly with PML lesion volume (r = 0.99, p = 0.009; see Fig, B). Data from our small cohort extend the findings of the study of Dalla Costa et al, demonstrating elevated serum NfL concentrations at PML onset. The 2 patients with the lowest NfL concentrations at PML onset and with the relatively smallest increase in NfL, either pre-PML in 1 patient or post-PML in the other patient, had the smallest lesion volumes on MRI at PML onset. In 2 patients, the increase of serum NfL prior to PML diagnosis could implicate earlier diagnostics by MRI or cerebrospinal fluid analysis. Considering the strong correlation between PML lesion volume and serum NfL, the easily accessible serum biomarker NfL may in the future have additional value to neuroimaging in the early detection and monitoring of (asymptomatic) PML. Future studies should focus on the predictive role of NfL as biomarkers in NTZ-PML, including larger numbers of patients for whom longitudinally collected, pre-PML blood samples are available.