Investigation of the Neutrophil-to-lymphocyte Ratio and Platelet-to-lymphocyte Ratio in Patients with Neuromyelitis Optica Spectrum Disorders


 Background: This study aimed to explore the differences in the neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) in patients with neuromyelitis optica spectrum disorders (NMOSDs) as well as their relationship with the onset of the diseases.Methods: The clinical data, laboratory findings, and imaging data of patients with NMOSD admitted to Perking University Third Hospital from January 2015 to December 2020 were retrospectively analyzed. Routine blood tests of patients performed within one week of the appearance of new clinical symptoms or imaging lesions were collected to calculate the NLR and PLR. The routine blood test of the patients in remission was performed more than 6 months after the patients stopped hormone use. The NLR and PLR of patients were compared with those of 100 healthy subjects undergoing physical examinations.Results: A total of 55 patients with NMOSD were enrolled. 44 patients with NMOSD were followed up. In patients with NMOSD, the white blood cell (WBC) count, absolute neutrophil count, and NLR were significantly higher than those in patients in remission and the controls, while the absolute lymphocyte count was significantly lower than that in patients in remission and the controls. In patients with NMOSD in remission, there were no statistically significant differences in the WBC count, absolute neutrophil count, absolute lymphocyte count, or NLR compared with the controls. The PLR of patients with NMOSD in the attack stage was significantly higher than that of the controls, while the PLR of patients with NMOSD in remission was not significantly different from that of the attack stage and the controls. There were no statistically significant differences between APQ4 (+) and APQ4 (-) in patients with NMOSD at the attack stage in the WBC count, absolute neutrophil count, absolute lymphocyte count, platelet count, NLR or PLR. ROC analysis of NLR and PLR for the diagnosis of inflammatory changes in NMOSD at the attack stage and controls: The ROC curve was plotted using NLR and PLR as dependent variables. In patients with NMOSD, the AUC was 0.806 for NLR and 0.612 for PLR. ROC analysis of NLR and PLR for the diagnosis of inflammatory changes in NMOSD at the attack stage and remission stage. The AUC was 0.728 for NLR and 0.594 for PLR.Conclusion: Patients with NMOSD had significantly higher WBC counts, absolute neutrophil counts and NLRs, and elevated NLRs were correlated with inflammatory activity in NMOSD.


Background
Neuromyelitis optica spectrum disorders (NMOSDs) are immune-mediated demyelinating diseases of the central nervous system (CNS). They mainly affect the optic nerve and spinal cord but may also involve other parts, such as the brain, cerebellum, and brainstem, and most patients are characterized by recurrent episodes. [1] NMOSD is likely to be combined with other autoimmune diseases, and abnormal immune responses play an important role in the development of NMOSD.
A routine blood test is one of the most commonly used blood tests in clinical practice, including indicators such as WBCs, neutrophils, and platelets. These indicators can re ect the basic in ammatory changes in the body. The neutrophil-to-lymphocyte ratio (NLR) is a biomarker for many diseases, such as myocardial infarction [2] and chronic obstructive pulmonary disease, [3] and an indicator for assessing prognosis. Increased neutrophil counts are correlated with in ammatory responses triggered by the relapse of MS. [4] NLR is signi cantly increased in patients with MS, [5] another demyelinating disease of the CNS and is correlated with the activity of the disease in relapsing-remitting MS. [6,7] The platelet-to-lymphocyte ratio (PLR) is also correlated with some autoimmune diseases. [8,9] The combination of NLR and PLR can better re ect in ammatory changes in the body. NMOSD is an autoimmune disease, and there are few studies on the correlations between NMOSD and NLR [10] and PLR. The present study aimed to investigate the correlations among NLR, PLR, and the onset of disease in patients with NMOSD during clinical episodes.
1 Study Subjects

Subjects
The data of 55 patients with NMOSD admitted to Perking University Third Hospital from January 2015 to December 2020 were retrospectively analyzed. The inclusion criteria were as follows: (1) patients with NMOSD who met the NMOSD diagnostic criteria developed by the International NMO Diagnostic Panel in 2015. [1] (2) patients who did not use hormones one month before the onset or relapse of the disease; and (3) patients with complete clinical data. The exclusion criteria were as follows: (1) patients with severe hepatic and renal dysfunction; (2) patients with malignancy, trauma, or autoimmune diseases; (3) patients with recent fever and infection; and (4) patients who were pregnant.

Methods
Clinical data, laboratory ndings, and imaging data were collected from 55 patients with NMOSD. The clinical data included sex, age, duration of disease, number of relapses, and Expanded Disability Status Scale (EDSS) score. Laboratory tests included routine blood tests, routine urine and stool tests, hepatic and renal function tests, blood sedimentation, thyroid function and antibodies, aquaporin 4 (AQP4), tumor markers, and immune-related indicators, such as anti-nuclear antibodies (ANAs) and anti-extractable nuclear antigen antibodies (ENAs). The routine blood test of the patient performed within one week of the onset of new clinical symptoms or imaging lesions was collected as clinical data, and the NLR and PLR were calculated. Forty-four patients with NMOSD were followed up. The routine blood test of the patients in remission was performed more than 6 months after the patients stopped hormone use. The results of routine blood tests and information, such as sex and age, of 100 healthy individuals were also collected.

Statistical analysis
The SPSS 25.0 software package (SPSS Inc., Chicago, IL, USA) was used for data analysis. Normally distributed measurement data are expressed as the means ± standard deviations, and nonnormally distributed measurement data are expressed as the medians (P25, P75). The t-test was used for intergroup comparisons between the normally distributed measurement data. One-way ANOVA was adopted for comparisons of the nonnormally distributed measurement data. The chi-square test or Fisher's exact probability method was used for comparisons of countable data. The diagnostic value was assessed based on the area under the curve (AUC) of the receiver operating characteristic (ROC) curve, with AUC < 0.50 indicating no predictability, 0.50 ≤ AUC ≤ 0.70 indicating low predictability, 0.70 < AUC ≤ 0.90 indicating medium predictability, and AUC > 0.90 indicating high predictability. P < 0.05 was considered statistically signi cant.

General characteristics of patients with NMOSD:
There were 55 patients in the NMOSD group, including 18 males and 37 females. The average age was 48.71 ± 16.43  years, and 36 patients were AQP4 (+), with a positivity of 65.45%. There were 36 males and 64 females in the healthy control group, with an average age of 46.45 ± 15.71 (25-80) years. There were no signi cant differences in sex or age between the three groups. The number of episodes in patients with NMOSD was 2.34 ± 1.67, the median duration of disease was 2 (1, 5) years, and the EDSS score was 3.70 ± 1.58. There were 13 males and 31 females in 44 follow-up patients, with an average age of 49.16±16.40  years. The EDSS score was 2.28±1.34, and 30 patients were AQP4 (+). The EDSS scores of NMOSD patients in the remission stage were signi cantly lower than those of NMOSD patients in the attack stage. Among the 55 patients with NMOSD, 11 took azathioprine, and 5 took mycophenolate mofelil. Among the 44 follow-up patients with NMOSD, 16 received azathioprine, and 9 received mycophenolate mofelil. There were signi cant differences in the use of immunosuppressants between NMOSD episodes and remission (p=0.005).

Comparison of the in ammatory indicators between patients with NMOSD at the attack stage and remission stage:
In patients with NMOSD at the attack stage, the white blood cell (WBC) count, absolute neutrophil count, and NLR were signi cantly higher than those in patients in remission and the controls, while the absolute lymphocyte count was signi cantly lower than that in patients in remission and the controls. The differences were statistically signi cant. In patients with NMOSD in remission, there were no statistically signi cant differences in the WBC count, absolute neutrophil count, absolute lymphocyte count, or NLR compared with the controls. There were no statistically signi cant differences in the platelet count among patients with NMOSD at the attack stage, patients with NMOSD in remission, and controls. The PLR of patients with NMOSD in the attack stage was signi cantly higher than that of the controls, while the PLR of patients with NMOSD in remission was not signi cantly different from that of the attack stage and the controls. (as shown in Table 1). there were no statistically signi cant differences in the WBC count, absolute neutrophil count, absolute lymphocyte count, platelet count, NLR or PLR (as shown in Table 2). Table 2 Comparison of the general characteristics of AQP4(+) and AQP4(-) in patients with NMOSD The ROC curve was plotted using NLR and PLR as dependent variables. In patients with NMOSD, the AUC was 0.806 for NLR (P = 0.000, 95% CI: 0.735 to 0.876) and 0.612 for PLR (P = 0.022, 95% CI: 0.519 to 0.705). The optimal cutoff point was 0.727, the sensitivity was 72.7%, and the speci city was 75%, as shown in Figure 1. ROC analysis of NLR and PLR for the diagnosis of in ammatory changes in NMOSD at attack stage and remission stage the AUC was 0.728 for NLR (P = 0.000, 95% CI: 0.629 to 0.828) and 0.594 for PLR (P = 0.108, 95% CI: 0.482 to 0.706). The optimal cutoff point was 2.3045, the sensitivity was 72.7%, and the speci city was 63.6%, as shown in Figure 2.

Discussion
Neutrophils are the most abundant cell type in the innate immune system and are associated with the pathogenesis of NMOSD. [11,12] Studies on animal models of experimental autoimmune encephalomyelitis have shown that the reduction in neutrophils may delay disease progression. Neutrophils may play an important role in the pathogenesis of experimental autoimmune encephalomyelitis by producing cytokines, damaging the blood-brain barrier, and promoting in ammation in the brain parenchyma. [13] It is now well established that neutrophils are the main in ammatory cells involved in early pathogenesis in models of NMOSD. [14] In ammatory cells in ltrating the lesions in NMOSD mainly include neutrophils, macrophages, and eosinophils, with relatively few T lymphocytes. [11] During the pathogenesis of NMOSD, neutrophils are the rst in ammatory cells to enter the lesion; entering may occur within a few hours after the onset of in ammation. Within 12 hours after onset, perivascular complement activation and loss of AQP4 with early myelin loss occur. In the chronic phase (seven days), there were almost no neutrophils in the lesion, and macrophages dominated and in ltrated extensively into the white matter. [11,15] In the present study, the WBC counts and absolute neutrophil counts during episodes of NMOSD were signi cantly higher than those in remission of NMOSD and the control group, whereas in patients with NMOSD in remission, there were no statistically signi cant differences compared with the controls, suggesting that neutrophils play a role during episodes of NMOSD. This nding is consistent with the pathophysiological process of NMOSD. Therefore, treatment for NMOSD should block the persistent in ammation produced by the apoptosis of neutrophils, promote the apoptosis of neutrophils, enhance the regression of in ammation, inhibit excessive tissue injury, and avoid persistent chronic in ammation. [16] The NLR is a systemic in ammatory marker based on routine blood tests, which are low in cost and easy to obtain. The NLR can re ect both the increase in neutrophils and the decrease in lymphocytes during the in ammatory response. The NLR has been correlated not only with some autoimmune diseases [9,17] but also with neurological disorders, such as Parkinson's disease [18] and Alzheimer's disease. [19] Several studies have veri ed that the NLR is correlated with the prognosis of diseases such as rheumatoid arthritis, [9] cancer, [20,21] and stroke. [22] Studies concerning NLR and NMOSD are rarely reported. Neutrophils are involved in the pathogenesis of NMOSD, [11,13,14] and the NLR can re ect changes in neutrophils in vivo. The present study revealed that the NLR was signi cantly higher in patients experiencing episodes of NMOSD than in patients in remission and controls. The change in NLR was due to an increase in the absolute neutrophil count and a decrease in the absolute lymphocyte count. Thus, it was suggested that the NLR could be used as one of the peripheral inflammatory biomarkers of disease activity in patients with NMOSD. The NLR and absolute neutrophil counts of patients with NMOSD at the attack stage were signi cantly higher than those of patients in remission, indicating that the in ammatory response in NMOSD onset might be more intense than that in remission, which also veri ed the difference in pathogenesis between the two different stages. This result suggests that the index has a moderate prediction value, indicating that the NLR might be a useful additional peripheral inflammatory biomarker in assessing disease activity in patients with NMOSD.
PLR is gaining importance as an indicator of in ammation. PLR in combination with NLR can better re ect the changes in in ammation in vivo and avoid the in uence of other factors on the absolute counts of leukocyte subtypes. The present study revealed no statistically signi cant differences in the absolute platelet count between patients with NMOSD at the attack stage, remission stage and the controls, but the absolute lymphocyte count of patients experiencing episodes of NMOSD was signi cantly lower than that of patients in remission and the controls. The PLR of patients with NMOSD attack stage was signi cantly higher than that of the controls, which may make the decrease of absolute lymphocytes count of patients with NMOSD attack stage larger.
There was a signi cant decrease in absolute lymphocyte count during episodes in patients with NMOSD, but there was no consistent change in PLR. There was no signi cant difference in PLR between the NMOSD attack stage and remission stage, but P =0.058, which was close to statistical signi cance. This value might be correlated with the small number of patients with NMOSD, so it would be necessary to expand the number of cases in further studies. This study investigated the sensitivity and speci city of NLR and PLR to in ammatory changes in NMOSD patients by drawing ROC curves. It was found that the changes in NLR in patients with attack stage were statistically signi cant compared with those in remission stage and the controls and better re ected in ammatory changes than PLR.
Drug use in patients with NMOSD may affect the WBC count, absolute neutrophil count, and absolute lymphocyte count, thereby affecting NLR and PLR. Some NMOSD patients took immunosuppressants in the onset and remission stages, and the proportion of immunosuppressants in the remission stage was signi cantly higher than that in the onset stage, which may partially affect NLR and PLR.
In conclusion, NLR might signi cantly increase in patients with NMOSD during episodes, and the increase in NLR might be correlated with in ammatory activity in NMOSD. The NLR could be used as a peripheral inflammatory biomarker of disease activity during an episode in patients with NMOSD and has some clinical applications. NLR can be a useful complement to MRI and CSF in the diagnosis of MS disease and is easier to achieve than the latter. However, the present study was a single-center study, and the number of patients enrolled was relatively small. The patients were taking different medications before relapse and after relapse, which might affect the results and have certain limitations. Further expansion is needed to increase the number of patients.

Declarations
Ethics approval and consent to participate This study was reviewed and approved by the Peking University Third Hospital Medical Science Research Ethics Committee (Item NO. 2021190). And this Ethics Committee exempted the informed consent due to retrospective nature of this study. All methods were carried out in accordance with the Declaration of Helsinki.

Figure 1
Receiver operating characteristic (ROC) curves of NLR and PLR for the diagnosis of in ammatory changes in patients with NMOSD during relapse compared to healthy controls Figure 2