Monocyte-to-Lymphocyte Ratio Was an Independent Factor of the Severity of Spinal Tuberculosis

Purpose The purpose was to explore the relationship between monocyte-to-lymphocyte ratio (MLR) and the severity of spinal tuberculosis. Methods A total of 1,000 clinical cases were collected, including 496 cases of spinal tuberculosis and 504 cases of nonspinal tuberculosis. Laboratory blood results were collected, including C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), white blood cells (WBC), hemoglobin (HGB), platelets (PLT), neutrophil count, percentage of neutrophils, lymphocyte count, percentage of lymphocytes, monocyte count, percentage of monocytes, MLR, platelets -to- monocyte ratio (PMR), platelets -to- lymphocyte ratio (PLR), neutrophil -to- lymphocyte ratio (NLR), and platelets -to- neutrophil ratio (PNR). The statistical parameters analyzed by the Least Absolute Shrinkage and Selection Operator (LASSO) and receiver-operating characteristic (ROC) curves were used to construct the nomogram. The nomogram was assessed by C-index, calibration curve, ROC curve, and decision curve analysis (DCA) curve. Results The C-index of the nomogram in the training set and external validation set was 0.801 and 0.861, respectively. Similarly, AUC was 0.801 in the former and 0.861 in the latter. The net benefit of the former nomogram ranged from 0.1 to 0.95 and 0.02 to 0.99 in the latter nomogram. Furthermore, there was a correlation between MLR and the severity of spinal tuberculosis. Conclusion MLR was an independent factor in the diagnosis of spinal tuberculosis and was associated with the severity of spinal tuberculosis. Additionally, MLR may be a predictor of active spinal tuberculosis.


Introduction
Tuberculosis poses a threat to human health all over the world and a huge economic burden on the people [1]. The incidence of spinal tuberculosis remains high, accounting for about 1%-3% of all tuberculosis [2]. Spinal tuberculosis is the most common type of extrapulmonary tuberculosis, accounting for 50% of skeletal tuberculosis [3]. Various examination methods have different limitations for the diagnosis of tuberculosis. The positive rates of acid-fast staining of sputum smears are usually used to diagnose tuberculosis. However, tuberculosis is easily missed due to the low sensi-tivity of the positive rate of acid-fast bacilli [4]. Although the culture of Mycobacterium tuberculosis is the gold standard for the diagnosis of tuberculosis, it cannot be used as an early diagnosis due to the low detection rate and long bacterial culture time [5]. In addition, the positive rate of spinal tuberculosis bacterial culture is only 44.39% reported in the literature [6]. Despite the rapid development of genetic testing, the positive rate of Mycobacterium tuberculosis DNA is only 66.67% [4]. Spinal tuberculosis is often diagnosed by pathological examination, but the sensitivity of histopathology of spinal tuberculosis ranges from 40.8% to 56.5% [7]. MLR has been reported in the literature as a predictor of tuberculosis diagnosis [8]. However, whether MLR can be used as a predictor in the diagnosis of spinal tuberculosis has not been reported.
MLR is closely associated with the expression of inflammatory factors in many diseases. Mehmet et al.'s study showed that MLR was not only positively correlated with microalbuminuria but also was determined as an independent factor of diabetic kidney injury by logistic regression analysis [9]. Aktas et al. found that MLR was involved in the inflammatory process of hepatic steatosis, and MLR was significantly higher in the hepatic steatosis group than in the control group [10]. MLR was significantly higher in irritable bowel syndrome, which may be a new inflammatory marker for irritable bowel syndrome diagnosed by laboratory examination [11]. In addition, MLR as an inflammatory marker was significantly different between colorectal cancer with and without metastasis, which may predict colorectal cancer metastasis [12]. Interestingly, COVID-19 is widespread around the world, and MLR as a marker of hematologic inflammation is a predictor involving diagnosis and clinical prognosis of the COVID-19 infection [13]. We hypothesized that MLR is also responsible for inflammatory factors in spinal tuberculosis.
ESR and CRP are usually highly expressed in inflammatory diseases and are used to assess the degree of infection of the disease [14]. Increasing ESR during treatment indicates severe spinal tuberculosis and poor prognosis [15]. CRP is as high as 41.9 mg/L before surgery for severe spinal tuberculosis and ESR is as high as 51.4 mm/h [16]. MLR has been confirmed as an inflammation marker in tuberculosis, and high levels of MLR are associated with severe tuberculosis [17]. However, the findings of Van et al. did not support the hypothesis that MLR was associated with tuberculin skin test positivity [18]. There are still contradictions in the association between MLR and tuberculosis. Furthermore, the association between MLR and the severity of spinal tuberculosis has not been reported.
In this paper, a total of 1,000 clinical data are collected, including training set data and external validation set data. We mainly discussed the diagnosis of MLR in spinal tuberculosis and the relationship between MLR and the severity of spinal tuberculosis. Furthermore, we also explored MLR as a predictor of active spinal tuberculosis.   . T-test was used to compare the two sets of data. The one-way ANOVA test was used to compare multiple sets of data. LASSO regression selected parameters with nonzero coefficients to obtain optimal lambda. Then, the parameters with AUC>0.6 were selected by the ROC curve. The predictive ability and accuracy of the nomogram were assessed by the C-index, calibration curve, ROC curve, and DCA curve. The external verification method further evaluated the predictive ability of the nomogram. Correlation analysis of data with normal distribution was used Pearson test, while Spearman test was used for data with nonnormal distribution. All continuous variables were represented by mean ± standard deviation (SD). P less than 0.05 was defined as a statistical difference. These parameters were used to construct a nomogram. The predictive ability of the nomogram ranged from 0.01 to 0.99 in the training set ( Figure 8). Similarly, the nomograms in the external validation set had the same predictive ability ( Figure 9). The C-index in the training set was 0.801 and that was 0.861 in the external validation set. The observations in the calibration curve were highly similar to the ideal values, whether in the training set ( Figure 10(a)) or the external validation set ( Figure 10(b)). AUC was 0.801 in the training set ( Figure 10(c)) and that was 0.861 in the external validation set ( Figure 10(d)). The net benefit of this nomogram ranged from 0.1 to 0.95 in the training set ( Figure 10(e)). Similarly, the benefit rate was also very high in the external verification set, ranging from 0.02 to 0.99 ( Figure 10(f)).

MLR Was Closely Related to the Severity of Spinal
Tuberculosis. 496 cases of spinal tuberculosis were included. Cases were divided into high ODI group (score≥20) and low ODI group (score<20) according to the ODI score. There were statistically different parameters between the two groups, including age, ESR, HGB, percentage of neutrophils, lymphocyte count, percentage of lymphocytes, MLR, PLR, and NLR ( Table 2). The MLR of the high ODI group was significantly higher than that of the low ODI group (Figure 11(a)). Cases were divided into low VAS group (score≤7) and high VAS group (score>7) according to the VAS score. There were statistically different parameters between the two groups, including age, PLR, and MLR. The MLR of the high VAS group was significantly higher than that of the low VAS group (Figure 11(b)). Cases were divided into five grades according to the ASIA classification. Parameters that were statistically different in ASIA included age, ESR, percentage of monocytes, and MLR ( Figure 11(c)). MLR was related to ESR ( Figure 11(d)) and CRP (Figure 11(e)). ESR was higher in the high ODI score group than in the low ODI score group (Figure 12(a)). Similarly, the level of ESR was statistically different in ASIA (Figure 12(c)). However, the level of ESR was not statistically different between the high VAS score group and the low VAS score group (Figure 12(b)). In the statistical analysis of ODI 5 Oxidative Medicine and Cellular Longevity score, VAS score, and ASIA, there was no statistical difference in CRP level ( Figure 12).

MLR as a Predictor of Active Spinal Tuberculosis.
Active spinal tuberculosis was defined based on moderate to severe dysfunction. Parameters with statistical differences were included in the ROC curve analysis, including age, ESR, HGB, percentage of neutrophils, lymphocyte count, percentage of lymphocytes, MLR, PLR, and NLR. Parameters with AUC>0.6 were selected to construct the nomogram, including age, percentage of lymphocytes, MLR, and PLR (Figure 13(a)). The ability of the nomogram to predict active spinal tuberculosis ranged from 0.1 to 0.9 ( Figure 13(b)). The observed value and the ideal value were consistent in the calibration curve (Figure 13(c)). AUC was 0.69 by ROC curve analysis (Figure 13(d)).

Discussion
The incidence of spinal tuberculosis in newborns has increased from 14% to 45.2%, posing a huge threat to humans [1,19,20]. Despite the rapid advancement of modern experimental technology, the positive detection rate of spinal tuberculosis is still not high. Diagnosis of spine tuberculosis usually uses the following methods: acid-fast tuberculosis smear, acid-fast tuberculosis DNA test, culture of mycobacterium tuberculosis, immunohistochemistry. However, the diagnostic value of MLR in spinal tuberculosis has not been reported.
MLR is significantly expressed in many diseases. MLR was used as a predictor of prostate cancer diagnosis and the AUC was 0.852 [21]. MLR>0.35 was associated with the overall survival rate of glioma patients [22]. High expression of MLR was detected in inflammatory diseases. High MLR expression was associated with mortality in acute respiratory distress syndrome [23]. MLR was significantly higher in patients with advanced knee synovitis than in early patients [23]. Our study found that MLR in the training set was statistically different between spinal tuberculosis and nonspinal tuberculosis. MLR was used as a predictor of spinal tuberculosis diagnosis, and its AUC was 0.663 by ROC curve analysis. The results of Sukson et al. showed that MLR was used as a predictor in the diagnosis of tuberculous pleuritis and the AUC was 0.91 [24]. Monocytes promoted the expression level of immune checkpoints and lead to latent tuberculosis infection [25]. MLR may affect the immune response of tuberculosis patients during the process of mycobacterium tuberculosis replication and progression to tuberculosis [26][27][28].
The severity of spinal tuberculosis had been reported to be associated with vitamin D receptor (VDR), osteopontin (OPN), and bone morphogenetic protein-4 (BMP-4) gene

10
Oxidative Medicine and Cellular Longevity polymorphism [29]. Osteopontin mediated inflammation by regulating immune cells [30]. Our results found that MLR was associated with the severity of spinal tuberculosis. Osteopontin might have been involved in regulating MLR to promote the progression of spinal tuberculosis. A previous study also reported that the concentration of OPN in tuberculosis plasma was higher than that of healthy patients [31]. VDR was associated with the clinical severity of spinal tuberculosis [29]. It may be that VDR regulated the expres-sion of multiple genes in monocytes [32]. MLR was closely related to the severity of the disease by promoting the activation of macrophages [33]. The results of our study were consistent with the finding that MLR was a marker of the severity of spinal tuberculosis. ESR and CRP were used as inflammatory markers, usually used to assess the degree of inflammation of the disease [34]. In multidrug-resistant tuberculosis, ESR and CRP were as high as 62.03 mm/h and 83.24 mg/L,  respectively, which were significantly higher than the ESR (37.20 mm/h) and CRP (50.02 mg/L) levels in drugsusceptible tuberculosis [35]. The expression of CRP was significantly increased in patients with advanced tuberculosis, but it lacked value in the diagnosis of early tuberculosis [36][37][38]. However, ESR and CRP also increased significantly in community-acquired pneumonia [39]. This made the diagnosis of spinal tuberculosis combined with community-acquired pneumonia more difficult. Our research showed that ESR and CRP had little contribution to the diagnosis of spinal tuberculosis and the severity of the disease. Our findings provided a distinction between the diagnosis of spinal tuberculosis and communityacquired pneumonia. MLR had been reported as a marker for active diseases, such as systemic lupus erythematosus [40]. MRL had also been reported to increase expression levels in active gout [33]. Monocytes promoted the release of 13 Oxidative Medicine and Cellular Longevity inflammatory mediators after pathogen invasion, and then transform into macrophages to participate in the immune response [41]. Inflammation was associated with low lymphocyte count [42]. Therefore, the increase in monocyte count and the decrease in lymphocyte count in inflammatory diseases lead to the imbalance of the MLR. Our find-ings also found that increased MLR was associated with the diagnosis of spinal tuberculosis, clinical severity, and active spinal tuberculosis.
In addition, PLR is also a marker of inflammation and is significantly expressed in a variety of diseases. Mehmet et al.'s study showed that PLR as an inflammatory factor 14 Oxidative Medicine and Cellular Longevity was significantly higher in advanced fibrosis 119 (61-1547) % than in mild fibrosis 99.5 (36-259) %, with a significant difference between the two groups [43]. Hamdi et al.'s study showed that PLR was positively correlated with CRP and ESR in the thyroid conditions, which could predict the thyroid conditions with high uptake or normal uptake [44]. The expression of PLR in malignant thyroid nodules was higher than that in benign thyroid nodules, and there was a positive correlation with TSH [45]. In addition, PLR expression was significantly higher in diabetes mellitus than in healthy patients, and there was a positive correlation between PLR and HbA1c [46]. Similarly, we also found that PLR, as an inflammatory factor, was significantly higher in spinal tuberculosis than in nonspinal tuberculosis in this study (P < 0:001). However, the AUC of MLR (0.663) in diagnosing spinal tuberculosis was higher than that of PLR (0.642) in the training set. The former (0.737) is also higher than the latter (0.706) in the verification set.

Conclusion
MLR as an independent factor was associated with the diagnosis of spinal tuberculosis and the severity of spinal tuber-culosis. Furthermore, MLR may be a predictor of active spinal tuberculosis.

Data Availability
The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author/s.

16
Oxidative Medicine and Cellular Longevity

Ethical Approval
All procedures were performed in accordance with relevant guidelines. This paper has been approved by the ethics committee of The First Affiliated Hospital of Guangxi Medical University.

Consent
Written informed consent of patients has been obtained for this study. The patient provided consent for publication of the data and images.

Disclosure
Funding bodies had no role in the study design, collection, analysis, and interpretation of the data or in writing the manuscript.

Conflicts of Interest
All authors declare no conflicts of interest.

Authors' Contributions
Liyi Chen and Chong Liu wrote the article and prepared Figures 1-13 and Tables 1 and 2. Zhaoping Gan and Xinli Zhan contributed equally to this work. Zhaoping Gan should be considered co-corresponding author. Liyi Chen and Chong Liu contributed equally to this work and cofirst authors. All authors reviewed the article.