Neutrophil-lymphocyte ratio in Guillain-Barré syndrome: A prognostic biomarker of severe disease and mechanical ventilation in Bangladesh.

Abstract

Background and Aims:

In addition to cellular and humoral immunity, inflammatory markers play an important role in the pathogenesis of Guillain-Barré syndrome (GBS) and are used to predict prognosis in many autoimmune diseases. The aim of this study was to identify whether the neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR) and monocyte-lymphocyte ratio (MLR) in the early stages of GBS have prognostic value for severe disease, mechanical ventilation (MV) and poor long-term outcome.

Methods:

A prospective cohort study of 140 adult patients with GBS and 140 healthy controls (HC) was performed in Bangladesh during 2019‒2022. Clinicodemographic characteristics of the patients were recorded, and hematological parameters were measured using an automated hematology analyzer.

Results:

Median (IQR) patient age was 35 (44‒23) years; 71% were male; 88% were severely affected (GBS Disability Score> 3); 32% required MV. Patients had higher NLR than HC (P<0.0001). Among patients, elevated NLR was associated with severe GBS and MV (P=0.001 and <0.0001, respectively) and moderately positively correlated with poor outcome at 4 weeks (r=0.423). Multiple logistic regression revealed NLR was an independent risk factor for severe GBS (OR=5.2, 95% CI=1.6–17.4) and MV (OR=1.5 1.1–2.1). No significant association was observed between elevated NLR and long-term outcome of GBS. ROC curves revealed NLR cut-off values of ≥2.432 and ≥4.4423 predicted severe disease (sensitivity=71%, specificity=75%, AUC=0.750, 95% CI=0.651–0.849, P=0.001) and MV (sensitivity=65.9%, specificity=81.7%, AUC=0.804, 95% CI=0.724–0.884; P<0.001).

Interpretation:

The NLR in the early stage of GBS may represent an independent prognostic factor of severe GBS and requirement for MV.

Introduction

Guillain-Barré syndrome (GBS), an acute immune-mediated inflammatory disorder, is characterized by rapidly progressive bilateral weakness that eventually affects the cranial nerve and respiratory muscles. ¹ The pathogenetic mechanisms of this autoimmune condition remain elusive. In two-thirds of cases, the aberrant response is triggered by infectious agents—mostly Campylobacter jejuni, which aggravates a humoral response and production of cross-reactive antibodies by molecular mimicry.¹ However, dysregulated innate and cell-mediated immune responses are also involved in induction of autoimmunity against host tissues after bacterial infections.¹ In GBS, cardiovascular and respiratory complications are the leading cause of death.² Approximately 30% of patients require mechanical ventilation (MV)^{1; 2} and about 20% cannot walk independently at 6 months after onset. ¹

Inflammatory factors such as neutrophils, lymphocytes and monocytes play crucial roles in triggering innate immune responses during the induction phase of systemic inflammation; however, their roles in GBS have not been investigated comprehensively.³ Elevated neutrophils are observed in the feces of C. jejuni-infected patients and high numbers of neutrophils result in human colonocyte cytotoxicity by releasing chemokines, cytokines and adhesion molecules,⁴ which suggests neutrophils are involved in the pathogenesis of C. jejuni-mediated disorders such as GBS through blood-brain barrier disruption.⁵ High circulating monocyte counts at the early phases of Parkinson’s disease (PD) and multiple sclerosis (MS) have been associated with the clinical severity of these diseases.^{6; 7} Recent reports suggested, variables derived from routine blood count parameters could represent simple indices of immune function and have prognostic significance in neurological and inflammatory diseases such as Alzheimer’s disease,⁸ PD,⁹ MS,¹⁰ systemic lupus eythematousus (SLE),¹¹ diabetic retinopathy¹² and bacterial infections.¹³ A higher neutrophil-lymphocyte ratio (NLR) has been associated with MS disability and activity and suggested as a rapid, inexpensive inflammatory marker.¹⁰ Recent studies showed an abnormal NLR and monocyte-lymphocyte ratio (MLR) were associated with autoimmune diseases.^{14; 15} A higher NLR has been associated with disease activity in Primary Sjogren’s syndrome¹⁴ and an early stage NLR of 2.94 was found to have prognostic value for disease activity in SLE.¹⁶

The ability to predict severe disease, respiratory insufficiency and a poor prognosis in the early-stage of GBS is crucial to enable appropriate early intervention. Clinical studies have identified multiple prognostic factors for severe GBS and MV.¹⁷ Previous studies examined the association between the NLR and disease severity and respiratory failure in GBS; however, the prognostic significance of the NLR for long-term outcomes was not addressed.^{18; 19} In addition, one report found no association between the platelet-lymphocyte ratio (PLR) and disease activity and weakness among children with GBS treated with intravenous immunoglobulin (IVIg).²⁰

To ameliorate the morbidity and mortality of the heterogeneous disorder GBS, it is important to identify readily available markers and clinical predictors to assess the risk of severe disease and poor prognosis at an early stage to allow individual patients to receive appropriate medical interventions. Thus, this study aimed to identify whether the NLR, PLR and MLR at onset can predict severe disease, MV and poor long-term outcome in GBS.

Methods and Materials

Study design and participants

This hospital-based prospective cohort study was conducted in the Gut-Brain Signaling Laboratory, icddr,b, Dhaka, Bangladesh, for 140 patients with GBS who satisfied the National Institute of Neurological and Communicative Disorders and Stroke (NINDS) criteria and were admitted to the National Institute of Neurosciences and Hospital (NINS), Dhaka, Bangladesh, between October 2019 and August 2022.²¹ All patients with GBS were enrolled within two weeks (14 days) of the onset of weakness and were aged ≥ 18 years.²¹ There were no exclusion criteria based on gender or ethnicity. Patients diagnosed with Miller Fisher syndrome, chronic inflammatory demyelinating polyradiculoneuropathy, spinal cord disease, pregnancy or other autoimmune diseases, neuropathies or neurological diseases were excluded. Demographic information including age, sex, height and weight were collected from the patient records and body mass index (BMI) was calculated. The clinical characteristics of the patients with GBS were recorded using a case report form, including the onset of weakness, antecedent infections, duration of hospitalization, autonomic dysfunction, sensory disturbances, hyporeflexia or areflexia, cranial nerves involvement, subtype of GBS based on electrophysiology tests, muscle strength, functional status and specific treatment.

A total of 140 healthy controls (HC) were recruited from people accompanying patients at NINS; the HC had no history of neurological complications or non-communicable diseases, and no recent history of infectious disease, pregnancy or surgery. Demographic information was recorded for all HC.

This study was reviewed and approved by the Institutional Review Board (IRB), icddr,b, Bangladesh and written informed consent was obtained from all participants prior to enrollment in the study.

Assessment of GBS disease severity and mechanical ventilation

The functional disability of patients with GBS was assessed using the GBS disability score (GBS-DS), a widely accepted scoring system also known as the Hughes Functional Grading Scale (HFGS), and defined as follows:²² 0, healthy state; 1, minor symptoms and capable of running; 2, able to walk 5 meters or more without assistance but unable to run; 3, able to walk 5 m across an open space with assistance; 4, bedridden or chair bound; 5, requires mechanical ventilation; 6, dead. In addition, muscle strength was evaluated using the Medical Research Council (MRC) sum score, which is defined as the sum of the MRC scores for six muscles in the upper and lower limbs on both sides; the score ranges from 60 (normal) to 0 (quadriplegic).²³ A MRC sum score of 0‒20 is defined as severe muscle weakness; 21‒40 and 41‒60 are defined as moderate and mild muscle weakness, respectively.²⁴

Disease severity was assessed using the GBS-DS at enrolment; patients able to walk independently with a GBS-DS < 3 were considered mildly affected and patients unable to walk independently with a GBS-DS of ≥ 3, severely affected.²⁵ Short-term disease outcome was assessed using the GBS-DS at 4 weeks and long-term disease outcomes were assessed using the GBS-DS at 13 weeks and 26 weeks;²² a GBS-DS ≥ 3 at follow-up was defined as poor prognosis.²⁵ Patients with respiratory distress and other minor complications including atelectasis, severe bulbar dysfunction, unable to clear bronchial secretions or unable to cough received MV at the discretion of the consultant/neurologist at the hospital in charge of the patient.²⁶ A GBS-DS of 5 during or within 7 days of enrolment to the study was classified as MV.²⁴

Laboratory investigations

Blood samples were drawn at the time of study enrolment from both patients with GBS and HC to measure hematological parameters. Whole blood samples were drawn into anticoagulant tubes containing ethylenediaminetetraacetic acid (EDTA) to measure hemoglobin (Hb), red blood cell count (RBC), white blood cell count (WBC), neutrophil count, lymphocyte count, platelet count and monocyte count using an automated hematology analyzer (Sysmex XN-1000, System Corporation, Kobe, Japan). The NLR ratio was obtained by dividing the absolute neutrophil count by the absolute lymphocyte count. Similarly, the PLR and MLR were obtained by dividing the platelet count by the absolute lymphocyte count and the absolute monocyte count by the absolute lymphocyte count, respectively.

Statistical analysis

All analyses were performed using Statistical Package for the Social Sciences for Windows 20.0 (SPSS Inc., Chicago, IL, USA). The chi-square test was employed to explore the associations of clinical and demographical characteristics with disease severity. The normality of the distribution of continuous variables was tested using the Kolmogorov–Smirnov test. Normally distributed variables were summarized as mean and standard deviation (SD); non-normally distributed variables, as median with interquartile range (IQR). The Mann-Whitney U-test was used to compare the median values for laboratory parameters between patients with GBS and HC; patients with severe and mild GBS; and MV and non-MV patients with GBS. Spearman’s correlation was used to assess the correlations between the laboratory parameters and the GBS-DS at the time of enrolment and 4, 13 and 26 weeks after enrolment. Spearman’s rho coefficient was denoted by ‘rs’ on a range from –1 to +1; a rs of 0‒0.19 indicates a very weak relationship with the GBS-DS, 0.20–0.39; weak, 0.40‒0.59; moderate, 0.60‒0.79; strong and 0.80‒1.0, very strong.²⁷

Logistic (unadjusted) regression was performed to measure the individual associations of the laboratory parameters with disease severity and MV. Subsequently, multivariable (adjusted) logistic regression was applied to calculate the odds ratio and 95% CI as effect sizes after adjusting for the other covariates based on our conceptual framework of factors that influence the associations of the NLR, PLR and MLR with disease severity and MV. Receiver operating characteristic (ROC) curves were used to determine the cut-off values for the NLR, PLR and MLR associated with severe disease, MV and poor long-term outcome. Area under the curve (AUC) analysis was used to assess and compare the overall prognostic performance of the NLR, PLR and MLR. Survival analysis of the duration of hospitalization was performed using the Kaplan-Meier method for the predicted NLR and MLR cut-off values.

Results

Clinicodemographic characteristics and laboratory findings for patients with GBS

The median age (IQR) of the 140 patients with GBS was 35 (44‒23) years with a male (71%) predominance. The median (IQR) age of the 140 HC was 34 (42‒27) years. A total of 102 patients with GBS (73%) had antecedent events, mostly diarrhea (28%), followed by respiratory tract infections (10%). Moreover, 48 patients with GBS (32%) required MV within 1 week of the onset of the disease. The median (IQR) duration of weakness before hospital admission for the patients was 5 (7–3) days and the duration of hospital stay was 10 (18–8) days. Overall, 39 patients with GBS (27%) had autonomic dysfunction and 47 patients (34%) had cranial nerve involvement. Bulbar palsy was the most common cranial nerve involvement in patients with GBS (27%). At the time of enrolment, 64 patients with GBS (46%) had severe muscle weakness (MRC sum score of 0‒20), 60 patients (43%) had moderate muscle weakness and 16 (11%) had mild muscle weakness. In addition, 67 patients with GBS (48%) had the axonal variant of GBS and 58 patients (41%) received only supportive treatment (Table 1).

Table 1:

Clinicodemographic characteristics of patients with GBS in Bangladesh

Variables	Patients with GBS (n = 140)
Demography
Age1	35 (44–23)
Male	100 (71%)
BMI1	23.2 (20.3–26.0)
Clinical manifestation
Duration of hospitalization (days)1	10 (18–8)
Duration of weakness before admission (days)1	5 (7–3)
Antecedent event	102 (73%)
Diarrhea	39 (28%)
Respiratory tract infections	14 (10%)
Autonomic dysfunction	39 (27%)
Mechanical Ventilation	48 (32%)
Cranial nerve involvement	47 (34%)
Oculomotor nerve	3 (2%)
Facial weakness	14 (10%)
Bulbar palsy	38 (27%)
MRC Sum score
Severe muscle weakness (0–20)	64 (46%)
Moderate muscle weakness (21–40)	60 (43%)
Mild muscle weakness (41–60)	16 (11%)
GBS-Disability Score (GBS-DS)
Severely affected (GBS-DS≥ 3)	123 (88%)
Mildly affected (GBS-DS < 3)	17 (12%)
Pain	130 (93%)
GBS Subtypes
Axonal (AMAN and AMSAN)	67 (48%)
Demyelinating (AIDP)	33 (24%)
Treatments strategies
IVIg	53 (38%)
Plasmapheresis	29 (21%)
Supportive care	58 (41%)

The table represents the clinical and demographical characteristics among 140 patients with GBS in Bangladesh. GBS severity is assessed based on GBS-disability score; severely affected (GBS ≥ 3), mildly affected (GBS < 3) and MRC Sum score is subcategorized into three, 0–20, 21–40 and 41–60.

¹median (IQR); BMI, body mass index; MRC, medical research council; IVIg, intravenous immunoglobulin.

Risk factors associated with GBS severity and outcome

The hematological parameters of the patients with GBS and HC are compared in Table 2. The median WBC, neutrophil count, lymphocyte count, monocyte count, platelet count, NLR, PLR and MLR were significantly higher in patients with GBS compared to HC. Clinicodemographic features and laboratory parameters were also compared among subgroups of patients based on disease severity and MV (Table 3). The duration of hospitalization was significantly longer for patients with severe GBS than patients with mild GBS and for MV patients compared to non-MV patients (P=0.013 and <0.0001, respectively). Autonomic dysfunction was significantly more common among MV patients and poor outcome (at 26 weeks) compared to non-MV and good outcome patients respectively (52% vs. 15%, P<0.0001 and 58% vs. 21%, P<0.001). Cranial nerves involvement was significantly associated with severe GBS and MV (P=0.042 and 0.026) and bulbar palsy was associated with both severe GBS and MV (P=0.035 and 0.017). Severe muscle weakness (MRC sum score of 0‒20 during the acute stage of disease) was significantly associated with severe GBS (P<0.001), MV (P=0.012) and poor outcome (P=<0.001). Significant number of severe GBS and MV patients received specific treatment (either IVIg or PE) (P=<0.001 and <0.001) compared to mild GBS and non-MV patients respectively (Table 3 and Supplementary Table 1).

Table 2:

Comparison of hematological parameters between the patients with GBS and healthy controls

Variables	GBS Patients (n = 140)	Healthy Controls (n = 140)	P-value
Laboratory findings [median, (IQR)]
Hb1 (g/dL)	13.3 (14.6–11.8)	13.9 (14.9–12.5)	0.217
RBC1 (1012/L)	4.9 (5.5–4.4)	4.8 (5.2–4.4)	0.486
WBC1 (109/L)	10.2 (13.4–8.0)	7.6 (8.64–6.49)	<0.0001
Neutrophil1 (109/L)	6.9 (10.4–4.9)	4.3 (5.5–3.6)	<0.0001
Lymphocyte1 (109/L)	2.1 (2.7–1.5)	2.4 (2.9–2.2)	0.006
Monocyte1 (109/L)	0.5 (0.6–0.4)	0.3 (0.4–0.3)	<0.0001
PLT1 (109/L)	299 (376.0–246.0)	258 (306.8–208.3)	0.002
NLR1	3.3 (5.4–2.2)	1.8 (2.1–1.5)	<0.0001
PLR1	143.7 (192.1–107.0)	107.5 (125.5–83.0)	<0.0001
MLR1	0.2 (0.3–0.2)	0.1 (0.2–0.1)	<0.0001

This table represents comparison of some hematological parameters between GBS patients and healthy controls.

¹median (IQR), Hb, hemoglobin; RBC, red blood cell count; WBC, white blood cell; PLT, platelet count; NLR, neutrophil-lymphocyte ratio; PLR, platelet-lymphocyte ratio; MLR, monocyte-lymphocyte ratio.

Table 3:

Risk factors associated with severe GBS and mechanical ventilation

Variables	Disease Severity		Pa value	Mechanical Ventilation		Pb value
	Severe (n= 123)	Mild (n = 17)		MV (n = 48)	Non-MV (n = 92)
Age1	35 (42–22)	41 (54–26)	0.234	34 (45–19)	34 (43–23)	0.634
Male	88 (72%)	12 (71%)	0.935	36 (75%)	64 (70%)	0.499
BMI1	23.2 (26.4–20.0)	23.4 (24.3–22.1)	0.678	23.6 (27.8–19.3)	23.1 (25.2–20.7)	0.537
Hospitalization in days1	10 (20–8)	9 (11–6)	0.013	25 (34–18)	9 (10–8)	<0.0001
Duration of weakness before hospital admission (days)1	5 (7–3)	6 (7.5–3)	0.811	4 (7–2)	5 (7–3)	0.058
Preceding events	91 (74%)	11 (65%)	0.420	36 (75%)	66 (72%)	0.680
Diarrhea	33 (27%)	6 (35%)	0.466	13 (27%)	26 (28%)	0.883
Autonomic dysfunction	36 (29%)	3 (18%)	0.316	25 (52%)	14 (15%)	<0.0001
Cranial nerve involvement	45 (37%)	2 (12%)	0.042	22 (46%)	25 (27%)	0.026
Facial weakness	12 (10%)	2 (12%)	0.796	5 (10%)	9 (10%)	0.906
Bulbar palsy	37 (30)	1 (6%)	0.035	19 (40%)	19 (21%)	0.017
Severe muscle weakness	64 (52%)	0	<0.001	39 (81%)	25 (27%)	0.012
Moderate muscle weakness	52 (42%)	8 (47%)		8 (17%)	52 (57%)
Mild muscle weakness	7 (6%)	9 (53%)		1 (2%)	15 (16%)
Treated (IVIg or PE)	81 (66%)	1 (6%)	<0.001	41 (85%)	41 (45%)	<0.001
Supportive care	42 (34%)	16 (94%)		7 (15%)	51 (55%)
WBC1 (109/L)	10.6 (13.7–8.4)	9.3 (9.5–7.2)	0.038	13.4 (16–9.6)	9.3 (11.3–7.0)	<0.0001
NLR1	3.6 (5.9–2.3)	2.2 (2.9–1.5)	0.001	6.1 (9.6–3.6)	2.5(3.8–1.6)	<0.0001
PLR1	143.8 (195.5–111.2)	127.6 (168.6–86.4)	0.130	163.0 (204.6–114.7)	134.3 (184.5–105.1)	0.099
MLR1	0.2 (0.3–0.2)	0.2 (0.3–0.1)	0.114	0.3 (0.5–0.2)	0.2 (0.3–0.1)	<0.0001

This table represents chi-square test (χ² test) applied on multiple variables to identify risk factors for disease severity and mechanical ventilation in patients with GBS.

¹median (IQR);

P^a, probability value between severe and mild GBS;

P^b, probability value between MV and Non-MV GBS; BMI, body mass index; MV, mechanical ventilation; MRC sum score, Medical Research Council-sum score at enrolment was used to categorize the patients based on their muscle weakness; Severe weakness (MRC sum score 0–20); Moderate muscle weakness (21–40); Mild muscle weakness (41–60); Based on treatment strategies, patients were categorized into “treated” and “Supportive care” groups; IVIg; intravenous immunoglobulin G and PE; plasma exchange; WBC, white blood cell; NLR, neutrophil-lymphocyte ratio; PLR, platelet-lymphocyte ratio; MLR, monocyte-lymphocyte ratio.

In addition, we determined the correlation between laboratory markers and GBS-DS at different follow-up timepoints of the disease. The WBC appeared to be correlated with the GBS-DS, with highly significant values at all follow-up timepoints. An elevated NLR and MLR levels were moderately positively correlated with the GBS-DS at the time of enrolment (r_s, P=0.538, <0.001 and r_s, P=0.417, <0.001) and weakly correlated with GBS-DS at 26 weeks (r_s, P =0.344, 0.001 and r_s, P= 0.246, 0.018) (Supplementary Table 2). Therefore, we compare these laboratory markers among GBS severity, MV and long-term outcome at 26 weeks. An elevated WBC was associated with both severe GBS and MV. The median NLR was significantly higher for patients with severe GBS and MV patients (P=0.001 and <0.0001, respectively) compared to patients with mild disease and non-MV patients. The median MLR was only significantly higher among MV patients than non-MV patients (P<0.0001). However, no significant differences in the median PLR were observed between these subgroups of patients (Table 3). Furthermore, higher WBC, NLR and MLR were identified as risk factors significantly associated with a poor outcome at 26 weeks (Supplementary Table 1).

Associations of NLR and MLR with disease severity and MV in GBS

Regression analysis (both un-adjusted and adjusted) was performed among the factors associated with severe GBS and MV (Table 4). In binary logistic regression in un-adjusted model showed, cranial nerve involvement (OR=4.3, 95% CI=0.9–19.8, P=0.049) and low MRC sum score (OR=18.6, 95% CI=5.5–63.2, P<0.001) were associated clinical risk factors for severe GBS. A high NLR was associated with a 1.8-times higher risk of severe disease (95% CI=1.2‒2.8, P=0.010) and 1.5-times higher risk of MV (95% CI=1.2‒1.7, P<0.001). Moreover, cranial nerve involvement (OR=2.3, 95% CI=1.1‒4.7, P=0.028), autonomic dysfunction (OR=6.1, 95% CI=2.7‒13.5, P<0.001), low MRC sum score (OR=9.2, 95% CI=1.2–71.6, P=0.035), WBC (OR=1.3, 95% CI=1.2‒1.4, P<0.001), PLR (OR=1.0, 95% CI=1.0‒1.1, P=0.019) and MLR (OR=35.0, 95% CI=4.0‒310.1, P=0.001) were also significantly associated with a higher risk of MV in the un-adjusted model (Table 4).

Table 4:

Logistics regression to identify risk factors for disease severity and mechanical ventilation in patients with GBS

Variables	Severe affected GBS				Mechanical ventilation
	Un-adjusted OR (95% CI)	P-value	Adjusted OR (95% CI)	P-value	Not-adjusted OR (95% CI)	P-value	Adjusted OR (95% CI)	P-value
Age	0.9 (0.9–1.0)	0.391	0.9 (0.9–1.0)	0.031	1.0 (0.9–1.0)	0.648	1.0 (0.9–1.0)	0.044
Sex	1.0 (0.3–3.2)	0.935	0.5 (0.1–3.3)	0.488	0.8 (0.4–1.7)	0.500	1.2 (0.4–3.8)	0.694
BMI	0.9 (0.9–1.1)	0.693	1.1 (0.9–1.4)	0.263	1.0 (1.0–1.1)	0.396	1.1 (1.0–1.2)	0.244
Preceding diarrhea	0.7 (0.2–2.0)	0.468	0.9 (0.2–5.1)	0.962	0.9 (0.4–2.1)	0.883	1.0 (0.3–3.3)	0.979
CNI	4.3 (0.9–19.8)	0.059	17.4 (1.2–251.4)	0.036	2.3 (1.1–4.7)	0.028	2.9 (1.1–8.3)	0.040
AD	1.9 (0.5–7.1)	0.323	0.3 (0.0–3.0)	0.325	6.1 (2.7–13.5)	<0.001	5.1 (1.6–16.1)	0.006
MRC Sum score	18.6 (5.5–63.2)	<0.001	33.0 (4.7–233.8)	<0.001	9.2 (1.2–71.6)	0.035	4.6 (0.5–42.3)	0.180
Weakness Duration before admission	1.0 (0.8–1.2)	0.991	1.2 (0.9–1.7)	0.160	0.9 (0.8–1.0)	0.052	0.9 (0.7–1.0)	0.147
WBC	1.2 (0.9–1.4)	0.081	0.8 (0.6–1.1)	0.166	1.3 (1.2–1.4)	<0.001	1.0 (0.8–1.2)	0.837
NLR	1. 8 (1.2–2.8)	0.010	5.3 (1.6–17.0)	0.005	1.5 (1.2–1.7)	<0.001	1.5 (1.1–2.2)	0.015
PLR	1.0 (0.9–1.0)	0.123	1.0 (0.9–1.0)	0.140	1.0 (1.0–1.1)	0.019	1.0 (0.9–1.0)	0.138
MLR	2.3 (0.1–37.8)	0.554	0.5 (0.0–53.3)	0.750	35.0 (4.0–310.1)	0.001	3.7 (0.1–101.1)	0.434

This table represents logistic regression applied on multiple variables with un-adjusted (univariate analysis) and adjusted (multivariate analysis) model to identify risk factors for disease severity and mechanical ventilation in patients with GBS. NLR is the only marker which has a significant association as a risk factor in both severe affected GBS and mechanical ventilation.

OR, odds ratio; 95% CI, 95% confidential interval; BMI, body mass index; CNI, cranial nerve involvement; AD, autonomic dysfunction; MRC sum score, Medical Research Council-sum score at enrolment; categorized into “Mild muscle weakness (MRC >40)” Severe-moderate muscle weakness (MRC≤40)”, where “Mild muscle weakness” was considered as reference group while multivariate analysis; WBC, white blood cell; NLR, neutrophil-lymphocyte ratio; PLR, platelet-lymphocyte ratio; MLR, monocyte-lymphocyte ratio.

Subsequently, multivariable logistic regression analysis was employed to evaluate the effect of associated factors on each other in the adjusted model with disease severity and requirement for MV in patients with GBS. After adjusting for age, sex, BMI, preceding diarrhea, cranial nerve involvement, autonomic dysfunction, MRC sum score, duration of weakness before admission, WBC, PLR and MLR, the NLR remained an independent risk factor for severe GBS and was associated with a 5.3-times higher risk of severe disease (95% CI=1.6‒17.0, P=0.005) and 1.5-times higher risk of MV (95% CI=1.1‒2.2, P=0.015). Moreover, older age, cranial nerve involvement and low MRC score were independent risk factor for severe GBS in the adjusted model. However, older age, autonomic dysfunction and cranial nerve involvement remained as independent risk factors for MV after adjusting for the other variables (Table 4).

Furthermore, analysis of the long-term disease outcome revealed duration of hospitalization, preceding diarrhea, autonomic dysfunction, low MRC sum score, WBC, NLR and MLR were risk factors associated with a poor outcome at 26 weeks in the un-adjusted model. However, after adjusting for other variables, only low MRC sum score and high MLR remained independent risk factors for a poor outcome at 26 weeks (Supplementary Table 3). No association was found between NLR and long-term poor outcome.

Clinical prognostic value of NLR, PLR and MLR for severe disease and MV in GBS

ROC curve analysis was used to identify the cut-off values and assess the prognostic sensitivity and specificity of the NLR, PLR and MLR for severe disease and MV in patients with GBS. The optimal cut-off value for NLR was 2.432, which had 71% sensitivity and 75% specificity to predict severe GBS. The AUC of NLR as a predictor for severe disease was 0.750 (95% CI: 0.65‒0.85, P=0.001), which was significantly higher than the AUC of the PLR (AUC: 0.617; 95% CI=0.48‒0.75; P=0.130) or MLR (AUC: 0.622; 95% CI=0.47‒0.78; P=0.114; Fig. 1A). A NLR cut-off value of 2.432 had a positive predictive value (PPV) of 95.6% and negative predictive value (NPV) of 30% for severe disease. We also determined the optimal prognostic cut-off values of the NLR and MLR for MV with consideration of the association observed with MV (Table 3). The ROC curve indicated the optimal NLR cut-off value for prediction of MV in patients with GBS was 4.4423 and this cut-off had 65.9% sensitivity, 81.7% specificity, an AUC of 0.804 (95% CI=0.72‒0.88; P=<0.001), PPV of 69.6% and NPV of 83.5%. Similarly, the optimal cut-off value for MLR to predict MV was 0.194 (67.8% sensitivity, 62.5% specificity, AUC 0.752 [95% CI=0.67‒0.84; P<0.001], PPV 43.7%, NPV 87.5%) (Fig. 1B). Moreover, we determined the optimal prognostic cut-off values of MLR for a poor long-term outcome at 26 weeks (based on Supplementary Table 1). The optimal cut-off value for MLR for a long-term poor prognosis was 0.3033 (58.3% sensitivity, 83.8% specificity, AUC 0.707 AUC [95% CI=0.58‒0.84; P=0.002], PPV 53.8%, NPV 85.9%; Fig. 1C).

Figure 1:

Receiver operating characteristic curve (ROC) analysis and corresponding areas under the curves (AUCs) of the prognostic value of the NLR, PLR and MLR in patients with GBS for: (A) severe disease—the optimal cut-off value for NLR was 2.432; (B) the need for MV—the optimal cut-offs for NLR, and MLR were 4.4423, and 0.194, respectively; and (C) poor long-term outcome—the optimal cut-offs for MLR was 0.3033.

We subcategorized the study patients based on the above-mentioned NLR (2.432 and 4.4423) and MLR (0.194 and 0.3033) cut-off values. NLR higher than 2.432 was significantly associated with severe GBS (P<0.0001; Fig. 2A). A NLR higher than 4.4423 was significantly associated with MV in patients with GBS (P<0.0001; Fig. 2B). A MLR lower than 0.194 was associated with non-MV patients with GBS (P<0.0001; Fig. 2C). Further analysis with long-term outcome at 26 weeks revealed, MLR level less than their optimum cut-off (0.3033) was significantly associated with a good outcome at 26 weeks (Fig. 2D).

Figure 2:

(A, B, C) Distribution of patients with GBS with severe disease (A), and who required MV (B), based on the identified NLR cut-off values of 2.432 and 4.4423. (C, D) Distribution of patients with GBS requiring MV (C) and with poor long-term outcomes (D) based on the MLR cut-offs of 0.194 and 0.3033, respectively.

Survival analysis of the associations with the duration of hospitalization

We categorized the patients with GBS into subgroups based on the optimal prognostic cut-off values for the NLR (2.432 and 4.4423) and MLR (0.194 and 0.3033). Survival analysis was performed to compare the duration of hospitalization for these subgroups of patients. Kaplan-Meier analysis showed the groups of patients with NLR ≥2.432 and ≥4.4423 had significantly longer durations of hospitalization compared to patients with NLR <2.432 and <4.4423 (log-rank test P=0.0007 and P=0.0001; Fig. 3A and 3B). Similarly, patients with MLR ≥0.194 and ≥0.3033 had longer durations of hospitalization than patients with MLR <0.194 and <0.3033 (log-rank test, P=0.0006 and P=0.0055; Fig. 3C and 3D).

Figure 3:

Survival analysis of the duration of hospitalization based on the A) NLR cut-off of 2.432, B) NLR cut-off of 4.4423, C) MLR cut-off of 0.194 and D) MLR cut-off of 0.3033.

Discussion

The NLR is considered to be an effective marker of systematic inflammatory responses in many autoimmune diseases. In this study, we investigated early clinical and inflammatory markers as prognostic factors for severe disease, MV and long-term poor outcome in patients with GBS. We found patients with GBS had higher NLR than the HC and that a high NLR at the early stage of GBS was an independent risk factor of severe disease and MV. Older age, cranial nerve involvement, autonomic dysfunction and low MRC sum score were identified as independent factors for severe GBS and MV. The optimum NLR cut-off values of 2.432 and 4.4423 at disease onset were significantly associated with severe GBS and MV.

Patients with GBS were previously shown to have a higher NLR compared to HC. ¹⁸ This finding was confirmed by our study and suggests systemic inflammation in patients with GBS leads to the initiation of immune tolerance breakdown after an infection. A high neutrophil count enhances the activation of T cells²⁸ and stimulates release of proinflammatory cytokines²⁹ and reactive oxidative species³⁰ that contribute to impairment of the blood brain barrier and trigger peripheral nerve inflammation. Moreover, a lower NLR is associated with reducing the development of autoimmune disease following specific treatment, which is indicating the involvement of NLR in the disease pathogenesis. ³¹

Clinicians often face difficulties in early recognition of patients with GBS who are likely to have a poor prognosis and thus require early intervention, and delays in treatment may lead to the occurrence of multiple complications that increase the risk of morbidity and mortality. Identification of factors that predict severe disease and poor short- and long-term outcomes is of major interest to many researchers. A large variety of studies have been conducted to investigate prognostic factors that may help to determine the disease severity and need for MV at the early stage of GBS.^{32; 33} In accordance with the results of this study, several study were reported autonomic dysfunction as an independent clinical predictor for MV patients with severe GBS.^{24; 32} The other predictors of severe GBS and MV were lower MRC sum score and cranial nerve involvement, which were confirmed by several studies including this study.^{24; 34; 35} Not every patients receive the specific treatment (IVIg or PE) due to the high-priced and no coverage of health insurance from the Government of Bangladesh. We found severe cases with GBS mostly received specific treatment in Bangladesh. Therefore, we excluded the treatment factor for predicting the disease severity and outcome to avoid the underestimation effect on the other factors in regression analysis. Moreover, multivariate analysis was performed to measure the effect of clinical risk factors on serum NLR levels, we found NLR as independent marker for severe GBS and MV. Furthermore, one study reported the NLR had higher prognostic accuracy than other inflammatory markers, such as C-reactive protein or neutrophil and lymphocyte counts.³⁶ A higher NLR has been proposed as an independent risk factor of patients with chronic respiratory failure and ventilator-associated pneumonia.^{37; 38} Our study confirms that the NLR represents an inflammatory prognostic factor for MV, which suggests an association exists between a high NLR and a higher risk of respiratory failure in GBS.

In agreement with our findings, severely affected patients were found to have a higher NLR compared to patients with mild GBS.¹⁸ A recent study determined a NLR cut-off of 3.05 was associated with severe functional disability in GBS and had a sensitivity of 67.3% and specificity of 67.3%.¹⁸ However, our study found the optimal cut-off value of the NLR for severe GBS was 2.432; this cut-off had a higher sensitivity (71%) and higher specificity (75%) for severe GBS. Another study reported a NLR cut-off of 4.76 was associated with the need for MV;³⁹ this value is very close to the NLR cut-off of 4.443 for MV determined in this study. Longer hospitalization was previously associated with MV.³⁹ Our study revealed elevated NLR and MLR were associated with longer durations of hospitalization in patients with GBS, which suggests higher NLR and MLR indicate the presence of more severe inflammation and a higher risk of serious illness.

A high NLR is an independent prognostic factor for poor outcomes in several diseases, including acute ischemic stroke, cardiac disorders and cancer.^40–42 Several studies reported that older age, severe deficits, dysautonomia, pulmonary complications, requirement of MV and lower MRC sum score at nadir were potential predictors of poor prognosis of GBS.^{43; 44} In this study, the median age of patients with GBS is lower than the western world. The age difference could partially be explained by the age distribution of the country and the number of people in each age group at risk of developing GBS in Bangladesh. In addition, our young adults are more frequently exposed to the Campylobacter-related infections and the infection ratio decreased with age.⁴⁵ Furthermore, we adjusted age with multiple risk factors (such as sex, BMI, preceding diarrhea, cranial nerve involvement, autonomic dysfunction, MRC sum score, weakness duration before admission and lab markers) and found increasing age as a risk factor for severe GBS and MV, but no association with the long-term outcome. Few studies have explored the value of NLR as a prognostic factor for severe disease and requirement for MV in GBS;^{18; 19} however, no previous studies investigated the association between the NLR and long-term outcome. This study examined both risk and prognostic factors associated with long-term disease outcome. Low MRC sum score and a higher MLR were found to be independent risk factors associated with poor long-term outcome at 26 weeks in Bangladesh. However, after adjusting for the effects of other cofactors, the NLR had no association with the long-term outcome in GBS.

The strength of this study is that the outcomes of the patients with GBS were clinically evaluated at four defined timepoints after enrolment. This prospective cohort study also obtained detailed information on involvement of the autonomic and cranial nervous system in patients with severe disease and who required MV; such data was lacking in previous studies.^{18; 19; 35} We identified few limitations in this study. NLR is considered as a potential marker for both systemic infections and autoimmune disorders. In GBS, inflammatory immune cells alter due to the peripheral nerve inflammation, and increased NLR could be one of them. It would have been provided more information as a precise marker of GBS if the data were compared with the C. jejuni enteritis cases without GBS as disease control. We did not systematically evaluated coexistence of other infections at enrollment or follow ups which might influence the NLR and outcome of the patients. Another limitation of this study is that we could not compare the changes in hematological parameters during the course of disease due to a lack of data.

We conclude that the NLR at the early stage of disease represents a novel, inexpensive and rapidly responding inflammatory biomarker of disease susceptibility, disease severity and the need for MV in GBS. Along with older age, cranial nerve involvement, autonomic dysfunction and low MRC sum score, the NLR may potentially provide an independent prognostic marker for severe GBS and requirement for MV, however, may not be effective in predicting long-term outcome in Bangladesh. We believe that inflammatory markers may play a role in early identification of a high risk of MV could enable early intervention measures and shorten the time to tracheotomy and ventilation, and thus reduce adverse complications for these patients with GBS. Further systemic investigations are required to evaluate the prognostic value of NLR and MLR in a large-cohort with infectious controls to include these biomarkers in the prognostic model for GBS.

Data availability statement

The data that supports the findings of this study is available upon request.