Peripheral immunity changes are associated with neurodegeneration and worse clinical outcome in idiopathic normal pressure hydrocephalus

Abstract

Background and purpose

Idiopathic normal pressure hydrocephalus (iNPH) pathogenesis is multifactorial. Systemic inflammation might have a role in gathering clinical–pathological trajectories. We aimed to shape the peripheral immune profile of iNPH and establish correlations with cerebrospinal fluid (CSF) markers, ventricular enlargement, and clinical outcomes.

Methods

We conducted a single‐center retrospective–longitudinal study, including 38 iNPH patients and 38 controls. Baseline iNPH Grading Scale and modified Rankin Scale (mRS) scores were collected with peripheral blood cell count, CSF amyloid‐β42 (Aβ42), total tau (t‐tau), phosphorylated‐181‐tau, and Evans index. Depending on 5‐year outcome, iNPH patients were grouped into “poor outcome” (PO; mRS ≥ 5) and “favorable outcome” (FO; mRS < 5). Biomarkers were compared and correlated with each other. Receiver operating characteristic analysis was performed.

Results

iNPH patients compared to controls had higher neutrophil‐to‐lymphocyte ratio (NLR; 2.43 ± 1.04 vs. 1.61 ± 0.47, p < 0.001), higher neutrophils (4.22 ± 0.86 1000/mL vs. 3.48 ± 1.34, p = 0.033), and lower lymphocytes (1.45 ± 0.55 1000/mL vs. 2.07 ± 0.86, p = 0.038), with the expected CSF biomarkers signature. In the patients' cohort, NLR was associated directly with t‐tau and inversely with Aβ42. NLR directly correlated with Evans index. PO patients compared to those with FO had higher NLR (3.25 ± 1.40 vs. 2.01 ± 0.77, p = 0.035) and higher t‐tau (274.76 ± 114.39 pg/mL vs. 150.28 ± 72.62, p = 0.017), with an area under the curve of 0.786 and 0.793, respectively.

Conclusions

iNPH patients present a proinflammatory state associated with neurodegeneration and predicting poor clinical outcome. Systemic inflammation represents a factor in the clinical–pathological progression of iNPH, and the NLR emerges as a potential prognostic index.

INTRODUCTION

Idiopathic normal pressure hydrocephalus (iNPH) is a common neurodegenerative disorder of the elderly [1] consisting of a heterogeneous syndrome (typically including motor and gait difficulties, cognitive deterioration, and sphincter incontinence), enlarged cerebral ventricles, and various neuropathological changes [2, 3].

The origin is multifactorial, with insufficient cerebral blood flow, impaired glymphatic circulation, and blood–brain barrier (BBB) disruption as the principal contributors to abnormal cerebrospinal fluid (CSF) dynamics and neurodegeneration. However, how these mechanisms interrelate remains to be elucidated [4].

Hypothetically, peripheral immunity and systemic inflammation might play a role in gathering pathogenic trajectories in iNPH, because they may affect BBB integrity and trigger substantial molecular changes at the brain level [5]. Although the presence of inflammation within the central nervous system has been demonstrated in iNPH, there is no proof, to date, on the peripheral events and their relationships with central neurodegeneration.

Peripheral blood cell count and the neutrophil‐to‐lymphocyte ratio (NLR) allow tracking the immune activation in the innate or adaptive branches [6]. An increased NLR consistent with a systemic proinflammatory state has been measured in many pathological conditions and neurodegenerative diseases, associated with a worse course, including mortality, and peculiar neuropathological signatures [7, 8, 9]. Conversely, there are still no data on iNPH.

Here, we conducted a longitudinal study to shape the peripheral immune profile of iNPH patients and dissect the brain–periphery interactions by examining correlations between immune parameters and CSF biomarkers of central neurodegeneration or clinical outcome measures.

METHODS

Study population

Thirty‐eight patients diagnosed with iNPH according to current guidelines and exhibiting classical neuroimaging features [10] were enrolled at the Neurology Unit of Tor Vergata University Hospital, Rome, Italy, during 2013–2023. Baseline scores of both the iNPH Grading Scale (iNPHGS) [11], which assesses the severity of gait, cognitive, and urinary dysfunction, and the modified Rankin Scale (mRS) [12], which measures general disability from absence (score = 0) to death (score = 6), were collected together with contemporary peripheral blood cell count, CSF biomarker levels, and the brain magnetic resonance imaging‐derived Evans index, which is defined as the ratio of the maximal width of the frontal horns to the maximum inner skull diameter [13]. Patients were prospectively observed and reassessed 5 years after baseline through the mRS score (available for 17 patients, none of whom developed features suggestive of a different diagnosis; other patients were lost to follow‐up). Depending on the follow‐up mRS score, patients were divided into the “poor outcome” (PO) group (mRS ≥ 5, including those with severe disability/bedridden or death) and the “favorable outcome” (FO) group (mRS < 5, including those without disability to moderate/requiring assistance for walking and daily living) [14, 15].

A control group of 38 sex‐/age‐matched participants, including n = 22 healthy volunteers and n = 16 non‐neurodegenerative patients receiving lumbar puncture for diagnostic purposes (finally diagnosed with headache, functional disorders, peripheral nervous system diseases) was also collected for comparative evaluation of blood cells count, CSF biomarkers, and Evans index calculation.

Subjects with main acute/chronic internal/inflammatory/infectious diseases potentially affecting blood counts, immunosuppressant/immunomodulant therapies, and other neurodegenerative diseases were excluded from the study.

Blood cell count

The blood was withdrawn in the morning, fasting (within 30 min from lumbar puncture if performed), and analyzed in the local laboratory, using an automated hematological analyzer (Dasit‐Sysmex, Milan, Italy). The leukocyte counts (neutrophils, lymphocytes, monocytes, eosinophils, and basophils) were recorded and the NLR calculated [7] for all patients with iNPH (n = 38) and all controls (n = 38).

CSF biomarkers assay

CSF was withdrawn by lumbar puncture. The levels of amyloid‐β42 (Aβ42), total tau (t‐tau), and phosphorylated‐181‐tau (p‐tau) were measured as previously described [16] in all iNPH patients (n = 38) and n = 16 controls.

Statistical analysis

Continuous variable distribution was preliminarily assessed by the Shapiro–Wilk test, and nonnormally distributed variables were eventually log10 + 1 transformed to allow for statistical calculation. Variables were compared between the groups through parametric (one‐way analysis of variance) or nonparametric (Mann–Whitney U) tests, as appropriate. Correlations were examined with Pearson test and further confirmed through a multiple linear regression analysis using age and sex as covariates. Categorical variables were compared through the chi‐squared test. Receiver operating characteristic (ROC) curve analysis was also performed, calculating the area under the curve (AUC) and cutoff points. Statistical analysis was run in blind using SPSS‐29 (IBM, Armonk, NY, USA). Significance was set at p < 0.05.

Standard protocol approvals, registrations, and patient consents

The study followed the ethical principles of Helsinki and was approved by the local Ethical Committee (protocol number 0026092/2017). All participants signed informed consent.

RESULTS

Clinical features and outcomes

Table 1 summarizes the main clinical–demographic parameters at baseline. Patients and controls were homogeneous in age and sex distribution.

TABLE 1.

Study population data.

Characteristic	iNPH, n = 38	CTLs, n = 38	Significance
Demographic parameters
Sex, M/F	19/19	18/20	n.s.
Age, years	74.13 ± 5.49	72.00 ± 5.90	n.s.
Shunt surgery, n	8	–	–
Clinical parameters at baseline
Disease duration, years	3.32 ± 2.26
iNPHGS, cognitive subscore	1.76 ± 0.90	–	–
iNPHGS, gait subscore	1.94 ± 0.92	–	–
iNPHGS, urinary subscore	1.75 ± 1.34	–	–
iNPHGS, total	5.50 ± 2.47	–	–
mRS	2.53 ± 0.80	–	–
Immunological parameters
WBC, 1000/mL	6.63 ± 1.71	6.46 ± 2.08	n.s.
Neutrophils, 1000/mL	4.04 ± 1.15	3.48 ± 1.34	p = 0.033
Lymphocytes, 1000/mL	1.89 ± 0.73	2.23 ± 0.79	p = 0.038
Monocytes, 1000/mL	0.51 ± 0.16	0.53 ± 0.17	n.s.
Eosinophils, 1000/mL	0.15 ± 0.09	0.17 ± 0.10	n.s.
Basophils, 1000/mL	0.02 ± 0.01	0.04 ± 0.02	n.s.
NLR	2.43 ± 1.04	1.61 ± 0.47	p < 0.001
CSF biomarkers, pg/mL
Aβ42 a	590.34 ± 309.90	962.30 ± 539.63	p = 0.002
t‐tau a	219.39 ± 100.24	249.86 ± 156.17	n.s.
p‐tau a	35.58 ± 21.20	37.12 ± 15.16	n.s.
Neuroradiological parameters
Evans index b	0.34 ± 0.04	0.27 ± 0.02	p < 0.001

Note: Values are given as mean ± SD. Statistical significance is indicated by bold font.

Abbreviations: Aβ42, amyloid‐β42; CSF, cerebrospinal fluid; CTL, control; F, female; iNPH, idiopathic normal pressure hydrocephalus; iNPHGS, iNPH Grading Scale; M, male; mRS, modified Rankin Scale; n.s., not significant; NLR, neutrophil‐to‐lymphocyte ratio; p‐tau, phosphorylated‐181‐tau; t‐tau, total tau; WBC, white blood cell count

^a Data are available for all iNPH subjects and 16 CTLs.

^b Data are available for 28 iNPH subjects and 10 CTLs.

Long‐term follow‐up (5 years) was available for 17 of 38 patients (45%). Of these 17 patients, eight (47%) underwent shunt surgery within 3 months after diagnosis (other patients refused or were not clinically indicated).

Seven patients (41%) had an FO (5‐year mRS < 5). Ten patients (59%) had a PO (5‐year mRS ≥ 5; in Table 2 main clinical–demographic parameters). Neither age/sex nor baseline clinical score differences were found between the groups. Nine of PO patients (five males and four females; two with surgery) were deceased (mRS = 6) within 5 years after diagnosis for cancer (n = 2), cardiovascular accidents (n = 4), and other nonspecified causes (n = 3). Therefore, we estimated that 53% of patients with iNPH (nine of the 17 longitudinally observed) died within 5 years from diagnosis, with cardiovascular accidents being the most common cause (44%).

TABLE 2.

Study iNPH cohort data.

Characteristic	PO, n = 10	FO, n = 7	Significance
Demographic parameters
Sex, M/F	05/05	05/02	n.s.
Age, years	74.10 ± 5.76	74.17 ± 4.35	n.s.
Shunt surgery, yes/no	2/8	6/1	p = 0.008
Clinical parameters at baseline
iNPHGS, cognitive subscore	2.10 ± 0.87	1.29 ± 0.75	n.s.
iNPHGS, gait subscore	2.00 ± 1.05	1.83 ± 0.75	n.s.
iNPHGS, urinary subscore	2.10 ± 1.49	1.14 ± 0.69	n.s.
iNPHGS, total	6.20 ± 2.82	4.33 ± 1.21	n.s.
mRS	2.70 ± 0.82	2.29 ± 0.75	n.s.
Immunological parameters
WBC, 1000/mL	6.30 ± 1.20	6.63 ± 2.36	n.s.
Neutrophils, 1000/mL	4.22 ± 0.86	3.91 ± 1.69	n.s.
Lymphocytes, 1000/mL	1.45 ± 0.55	2.07 ± 0.86	n.s.
Monocytes, 1000/mL	0.47 ± 0.17	0.45 ± 0.17	n.s.
Eosinophils, 1000/mL	0.12 ± 0.08	0.16 ± 0.15	n.s.
Basophils, 1000/mL	0.02 ± 0.01	0.03 ± 0.01	n.s.
NLR	3.25 ± 1.40	2.01 ± 0.77	p = 0.035
CSF biomarkers, pg/mL
Aβ42	673.29 ± 331.71	594.14 ± 192.51	n.s.
t‐tau	274.76 ± 114.39	150.28 ± 72.62	p = 0.017
p‐tau	38.58 ± 19.96	28.83 ± 16.42	n.s.
Neuroradiological parameters
Evans index	0.34 ± 0.02	0.34 ± 0.07	n.s.

Note: Values are given as mean ± SD. Statistical significance is indicated by bold font.

Abbreviations: Aβ42, amyloid‐β42; CSF, cerebrospinal fluid; F, female; FO, favorable outcome; iNPH, idiopathic normal pressure hydrocephalus; iNPHGS, iNPH Grading Scale; M, male; mRS, modified Rankin Scale; n.s., not significant; NLR, neutrophil‐to‐lymphocyte ratio; PO, poor outcome; p‐tau, phosphorylated‐181‐tau; t‐tau, total tau; WBC, white blood cell count.

Immune parameters, CSF biomarkers, and Evans index

Regarding the blood cell count, at baseline, iNPH patients compared to controls had higher NLR (F _{1, 68} = 20.37, p < 0.001), higher neutrophile count (F _{1, 68} = 4.71, p = 0.033), and lower lymphocytes count (F _{1, 68} = 4.47, p = 0.038). Other subpopulations did not differ between patients and controls.

Regarding CSF biomarkers, at baseline, patients with iNPH compared to controls had lower CSF Aβ42 levels (F _{1, 47} = 11.37, p = 0.002). CSF t‐tau (F _{1, 47} = 0.096, p = 0.759) and p‐tau (F _{1, 47} = 1.99, p = 0.164) levels were slightly reduced in patients with iNPH although not significantly.

As expected, the Evans index was significantly higher in iNPH than controls (F _{1, 36} = 27.03, p < 0.001).

In the iNPH group, but not in the control group, immunity parameters showed significant associations with CSF biomarkers. Specifically, iNPH participants showed a direct correlation between NLR and CSF t‐tau levels (r ₂₈ = 0.50, p = 0.004), and an inverse correlation between NLR and CSF Aβ42 levels (r ₂₈ = −0.49, p = 0.006); these correlations were both confirmed in a model adjusted for age (R ² = 0.29, F _{3, 26} = 3.62, p = 0.026) and sex (R ² = 0.30, F _{3, 26} = 3.77, p = 0.022), as seen in Figure 1. Lymphocyte count was negatively correlated with CSF t‐tau levels (r ₂₈ = 0.36, p = 0.047), although significance was lost in the adjusted model. Finally, in iNPH, NLR directly correlated with Evans index (r ₂₀ = 0.49, p = 0.021), although the significance was lost in the model adjusted for age and sex.

FIGURE 1.

Association of neutrophil‐to‐lymphocyte ratio (NLR) with cerebrospinal fluid (CSF) total tau (t‐tau) and amyloid‐β42 (Aβ42). Dots represent subjects in the study. Solid lines represent regression lines generated from all subjects combined. Dashed lines represent 95% confidence interval. **p ≤ 0.01. (a) Association between NLR and CSF t‐tau. (b) Association between NLR andCSF Aβ42.

Clinical correlations

A direct correlation was found between CSF t‐tau levels and the baseline iNPHGS cognitive subscore (r ₃₀ = 0.57, p = 0.016), albeit the significance was lost in the model adjusted for age and sex. No further correlations with baseline clinical scores emerged, neither for other CSF biomarkers nor for immune parameters.

Regarding the long‐term outcome, patients with PO had NLR values higher than those with FO (F _{1, 15} = 5.35, p = 0.035), as seen in Figure 2. The ROC analysis provided an AUC for NLR of 0.786 (p = 0.05); at the given cutoff value of 2.23, NLR differentiated PO versus FO patients with a sensitivity of 70% and a specificity of 71.4%, as seen in Figure 3.

FIGURE 2.

Scatter plots of neutrophil‐to‐lymphocyte ratio (NLR) values and cerebrospinal fluid (CSF) total tau (t‐tau) levels in favorable outcome (FO) and poor outcome (PO) groups, with median and 25th–75th interquartile range. *p ≤ 0.05. (a) NLR values comparison. Black dots represent patients from FO cohort; red dots represent patients from PO cohort. (b) CSF t‐tau levels. Black dots represent patients from FO cohort; red dots represent patients from PO cohort.

FIGURE 3.

Receiver operating characteristic (ROC) curves of neutrophil‐to‐lymphocyte ratio (NLR) and total tau (t‐tau). *p ≤ 0.05. (a) ROC curve of NLR to discriminate between poor outcome (PO) and favorable outcome (FO) groups. (b) ROC curve of cerebrospinal fluid t‐tau to discriminate between PO and FO groups.

CSF t‐tau levels were also significantly higher in patients with PO compared to those with FO (F _{1, 15} = 7.25, p = 0.017), as seen in Figure 2. The ROC analysis provided an AUC for CSF t‐tau levels of 0.793 (p = 0.045); at the given cutoff value of 165.0 pg/mL, CSF t‐tau differentiated PO versus FO patients with a sensitivity of 80.0% and a specificity of 60.0%, as seen in Figure 3.

Evans index did not correlate with clinical parameters or differ between the two groups.

DISCUSSION

This longitudinal study aimed at unraveling the contribution of peripheral immunity to iNPH clinical–pathological features. Of interest, we found that iNPH patients may present with a higher blood NLR (increased neutrophil and reduced lymphocyte counts), predictive of heavier central neurodegeneration and poorer long‐term clinical outcomes.

Changes in peripheral blood cell count in iNPH patients basically reflect a shift of the immune system toward a proinflammatory state, which is in line with findings from other neurodegenerative diseases, such as Alzheimer and Parkinson diseases [17, 18]. For instance, although evidence in these different conditions is more robust and putative mechanisms have been outlined, in iNPH, we can only suppose that comorbid vascular risk factors, which typically account for systemic inflammation, may affect the peripheral immune profile and trigger classical pathogenic events [19, 20]. On the other hand, the peripheral immunity modifications may follow the neuropathological process, resulting from the leukocyte brain infiltrates associated with neurodegeneration [21, 22], or propagation of central neuroinflammation due to the release of cytokines and chemokines [23].

Regardless, we noticed that, only in the iNPH group but not in controls, the NLR was independently associated with CSF t‐tau and Aβ42 levels, suggesting some relationships between peripheral and central events. Although a disease‐specific biomarker for iNPH is still lacking, the available CSF markers (Aβ42 peptide and tau proteins) allow tracking the single molecular pathways involved in the complex of neuropathology [16, 24, 25]. As expected, CSF Aβ42 levels were reduced in our cohort, whereas t‐tau and p‐tau were not different from the controls. Lower CSF Aβ42 is now considered a biochemical signature of iNPH [26], resulting from the downregulation of amyloid production due to periventricular hypometabolism and impaired interstitial fluid circulation within the brain [20, 25, 27]. The CSF t‐tau protein, instead, mostly reflects nonspecific neurodegeneration or axonal injury, and the levels can be similar to or lower than controls because of a potential diluting effect mediated by the excessive CSF volume [25, 27]. However, the increase of t‐tau levels in iNPH patients has often been associated with a worse phenotype, making CSF t‐tau a clinical–pathological aggressiveness biomarker [25, 28, 29]. Accordingly, here, the CSF t‐tau levels were, to some extent, correlated with greater clinical severity (cognitive iNPHGS subscore).

NLR also directly correlated with ventricular enlargement, measured through the Evans index, within a simple statistical model. Therefore, the associations of NLR with ventricular size and CSF levels of Aβ42 and t‐tau, in particular, may indicate that systemic inflammation accompanies or sustains the central neurodegeneration in iNPH, contributing to disease progression. In this regard, there are novel, supportive data showing how systemic inflammation can aggravate the glymphatic flow, BBB permeability, and CSF circulation [30], thus theoretically contributing to brain damage in iNPH.

In agreement with this hypothesis, we found that higher NLR and CSF t‐tau values at the diagnosis time can identify patients with worse long‐term clinical course.

Our longitudinal observation over 5 years showed that >50% of patients developed a high‐level disability or died, even when shunt surgery was performed. These results basically mirror larger epidemiological studies displaying a 5‐year mortality reaching almost 90% for untreated cases [31] and 40% for those treated [32] (with some differences due to shunt response), with cardiovascular causes [21] being the most prevalent. Therefore, in iNPH, prognostic biomarkers are particularly needed to improve and personalize treatments, avoiding risky, onerous, or unnecessary procedures.

NLR and CSF t‐tau may thus represent potential indexes to stratify iNPH patients from the early stages. The predictive significance of CSF t‐tau was already suggested by works demonstrating the poor shunt surgery responsiveness of patients with higher CSF t‐tau [28]. Conversely, the predictive value of immune parameters, namely the NLR, is unprecedented. The accuracy of NLR (64.7%, at the cutoff value of 2.23) to detect patients at risk for bad outcomes (severe disability or death) within 5 years after the diagnosis was lower than that of CSF t‐tau (70.5%, at the cutoff value of 165.0 pg/mL); however, the NLR is a more easily accessible and cheaper biomarker, with substantial potential in clinical practice.

This study has several limitations, including the sample size, the heterogeneity of the population, the bare clinical assessment, the exiguity of the CSF and blood biomarkers panel, the roughness of Evans index, and a potential attrition bias. Future works in this direction might take advantage of a deeper immunological assessment in peripheral blood mononuclear cells [33], a more detailed analysis of other neuroimaging markers [34, 35], and proper gait analysis with objective measurements.

Nevertheless, we demonstrated a role for systemic inflammation in iNPH, showing that some peripheral immune changes occur, in association with central neurodegeneration and clinical progression. In iNPH patients, the NLR correlates with CSF t‐tau levels, and both predict the long‐term clinical outcome, emerging as potential prognostic indexes. However, further confirmation in larger samples is now due.

Funding information

The research leading to these results has received funding from the European Union (NextGenerationEU) through the Italian Ministry of University and Research under PNRR–M4C2‐I1.3 Project PE_00000019 “Heal Italia” to T.S. (CUP E83C22004670001). The views and opinions expressed are those of the authors only and do not necessarily reflect those of the European Union or the European Commission.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

INFORMED CONSENT STATEMENT

Written informed consent was obtained from all the participants before the collection of clinical data and cerebrospinal fluid.

Bissacco J, Simonetta C, Mascioli D, et al. Peripheral immunity changes are associated with neurodegeneration and worse clinical outcome in idiopathic normal pressure hydrocephalus. Eur J Neurol. 2024;31:e16179. doi: 10.1111/ene.16179

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.