Neuronal Pentraxin 2 as a Potential Biomarker for Nusinersen Therapy Response in Adults with Spinal Muscular Atrophy: A Pilot Study

Abstract

Background: The treatment landscape for spinal muscular atrophy (SMA) has changed significantly with the approval of gene-based therapies such as nusinersen for adults with SMA (pwSMA). Despite their efficacy, high costs and treatment burden highlight the need for biomarkers to objectify or predict treatment response. This study aimed to identify such biomarkers. Methods: A proteomic analysis of cerebrospinal fluid (CSF) from pwSMA (n = 7), who either significantly improved (SMA Improvers) or did not improve in motor function (SMA Non-Improvers) under nusinersen therapy, was performed. Data are available via ProteomeXchange with identifier PXD065345. Candidate biomarkers-Neuronal Pentraxin 2 (NPTX2), Contactin 5 (CNTN5), and Anthrax Toxin Receptor 1 (ANTXR1)-were investigated by ELISA in serum and CSF from an independent pwSMA cohort (n = 14) at baseline, 2 and 14 months after therapy initiation. Biomarker concentrations were correlated with clinical outcomes. Additionally, NPTX2 was stained in spinal cord sections from a mild SMA mouse model (FVB.Cg-Smn1tm1Hung Tg(SMN2)2Hung/J). Results: CSF NPTX2 levels decreased in pwSMA after 14 months of nusinersen therapy, independent of clinical response. The change in NPTX2 serum levels over 14 months of nusinersen treatment correlated with the change in HFMSE during this period. CNTN5 and ANTXR1 showed no significant changes. In the SMA mouse model, NPTX2 immunoreactivity increased at motoneuron loss onset. Conclusions: NPTX2 emerges as a potential biomarker of treatment response to nusinersen in pwSMA suggesting its significant pathophysiological role in late-onset SMA, warranting further investigation.

Keywords: NPTX2; biomarker; nusinersen; spinal muscular atrophy.

Figure 1

Overview of the study design. Control samples were only available from a single time point. pwSMA = adult individuals with 5q-SMA; SMA Improvers = increase ≥ 3 points in HMFSE after 6 months of treatment with nusinersen; SMA Non-Improvers = increase of < 3 points in HFMSE after 6 months of treatment with nusinersen; HFMSE = Hammersmith Functional Motor Scale Expanded; RULM = Revised Upper Limb Module; mo. = months; wt = wild type; P20, P40, P72, and P>400 refer to postnatal days, with P20 indicating postnatal day 20.

Figure 2

Results of the proteomic analysis with (a) Heatmap of differentially expressed proteins between pwSMA and controls (Ctrl). The color scale on the right represents Z-score values, with yellow shades indicating increased expression levels and purple shades indicating decreased expression levels. SMA BS (Baseline) I to V represent SMA Improvers (HFMSE improvement ≥ 3 points after 6 months of nusinersen therapy), while SMA BS VI and VII represent SMA Non-Improvers (HFMSE improvement < 3 points). Each row corresponds to a protein and each column represents a sample. The visual cluster structure of the heatmap reveals differences in protein expression between individuals with SMA and controls; (b) Volcano Plot of differentially expressed proteins between pwSMA and controls (ctrl). The logarithmic change in expression levels (log2 fold change) is represented on the X-axis, while the statistical significance is shown as the negative decadic logarithm of the p-value (−log10(p-value)) on the Y-axis. Each point represents a different protein, with color-coded points indicating statistically significant differences (p < 0.05) between pwSMA and controls. Red points indicate a positive change in protein expression compared to controls, while blue points represent a negative change; (c) Results of proteomic profiling from CSF with baseline levels of CNTN5, ANTXR1, and NPTX2 in controls and SMA Improvers and Non-Improvers. Significant differences (p <0.05) are marked with an asterix (*). CNTN5 levels are decreased in SMA Improvers compared to Non-Improvers, with no significant difference compared to controls. ANTXR1 is elevated in SMA Improvers compared to Non-Improvers, again without a significant difference compared to controls. For NPTX2, no significant group differences were observed.

Figure 3

(a) CSF and (d) Serum levels of NPTX2 in controls and pwSMA at T0 (baseline), T1 (2 months after start of treatment nusinersen), and T2 (14 months after start of treatment with nusinersen) across all pwSMA shown in a bar plot with individual data points and error bars representing the standard deviation. The bar height represents the mean value. Across the entire pwSMA cohort, CSF NPTX2 levels were lower at T2 compared to T0 and lower at T2 compared to T1. No significant differences were observed in serum levels across time points. (b) CSF and (e) Serum levels of NPTX2 at T0, T1, and T2 in SMA Improvers (improvement in HFMSE ≥ 3 points after 6 months of nusinersen therapy) vs. SMA Non-improvers (improvement in HFMSE < 3 points after 6 months of nusinersen therapy). Significant differences (p < 0.05) are marked by an asterisk *. Note that the Y-axis scaling differs between serum and CSF values due to a different range of protein concentration. At T2, serum NPTX2 levels were lower in SMA Improvers compared to Non-Improvers. No significant differences in CSF NPTX2 levels were observed between SMA Improvers and Non-Improvers at the respective time points. Individual longitudinal trajectories of NPTX2 (c) CSF and (f) serum levels are shown in a line graph across all pwSMA, with SMA Improvers highlighted in green and SMA Non-Improvers in red.

Figure 4

Correlation of the change in NPTX2 serum concentration and HFMSE score from T0 (baseline) to T2 (14 months after treatment initiation with nusinersen) (r = −0.688, p = 0.019, n = 11). The points represent individual pwSMA. The line of best fit was calculated using simple linear regression, R² = 0.52, p = 0.01.

Figure 5

Immunostaining of (a) spinal motoneurons (SMI-32, red) and (b) NPTX2 (green) in the ventral horn of lumbar spinal cord slices from wild type (wt) and SMA mice (SMA) at P20 (postnatal day 20), P42 and P70 and P>400. The SMA values normalized to the wt values for the immunoreactivity of NPTX2 and the absolute number of motoneurons at the different time points are shown on the right (n = 9 for P20 and P>400; n = 7 for P42 and P70; three SMA mice per condition with two or three slices per animal). Scale bar: 20 µm. From P42 onwards, the number of motor neurons is reduced in SMA mice compared to wt mice. At P42–but not at any other time point–NPTX2 immunofluorescence is increased in SMA compared to wt mice. Significant differences (p < 0.05) are marked by an asterisk *. (c) Dynamics of wt-normalized values for motoneuron count and NPTX2 immunoreactivity in SMA mice across the different time points. In SMA mice, the number of motor neurons continues to decline progressively from P42 onwards, while NPTX2 immunofluorescence peaks at P42.