Platelet-To-Lymphocyte Ratio Can Be Used As Marker To Predict Activity And Severity In Moderate And High Disease Activity of Rheumatoid Arthritis: A Retrospective Study

Objective: This study aimed to study and evaluate the value of platelet-to-lymphocyte ratio (PLR) and neutrophil-to-lymphocyte ratio (NLR) as markers to predict the disease activity and severity in patients with rheumatoid arthritis (RA). Methods: A total of 139 inpatients with RA were divided into two groups: moderate activity group (3.2<DAS28-CRP score<5.1) and high activity group ( ≥ 5.1). The correlation of routine hematological indices with DAS28 and joint ultrasound score (gray-scale ultrasound score, GS-US; power Doppler ultrasound score, PD-US; total ultrasound score, T-US) were analyzed by Pearson's correlation and logistic regression analyses. Receiver operating characteristics (ROC) analysis was performed to compare the ecacy of blood indices, ESR, or CRP in reecting the disease activity and severity of RA. Results: The values of PLR, NLR, PD-US, and T-US were signicantly different between moderate and high disease activity groups (p<0.001), and PLR was signicantly correlated with PD-US and DAS28. Logistic regression analyses showed that PLR was an independent risk factor for disease activity by DAS 28 and joint damage severity by PD-US and T-US. ROC analysis showed that the ecacy of using PLR alone to evaluate the disease activity and joint severity of RA was similar to that of using combined CRP and ESR. The best cut-off value of PLR for predicting high disease activity and high joint severity was determined as 236.6. Conclusions: PLR can be used as a marker to predict activity and severity in patients with moderate and high RA disease activity. our knowledge, few comparative analyses have been conducted to evaluate the relationship between routine hematological indicators with joint ultrasound score and DAS28. Our study aimed to assess the correlations between the levels of routine hematological indices with different disease activities and joint ultrasound scores in patients with RA. The purpose was to explore if routine hematological indices can be used to assess disease activity and joint severity of RA in a cost-effective and convenient manner. DAS28 and (ROC) to compare the ecacy of blood indices, ESR, or CRP in reecting the disease activity and severity of RA. Values with two-sided p-values of less than 0.05 were considered statistically signicant. activity and joint severity of RA in a cost-effective and convenient manner. The results showed that PLR, NLR, PD-US, and T-US were signicantly different between moderate and high disease activity groups. PLR, a clinical index routinely measured in hematological tests, was signicantly correlated with PD-US and DAS28. Moreover, the results of logistic regression analysis showed that PLR is an independent risk factor for disease activity and joint severity in RA. The ecacy of using PLR alone to evaluate disease activity and severity was similar to that of using combined CRP and ESR. PLR is a blood parameter that can be economically and conveniently assessed during routine blood tests in developing countries, PLR can be a simple and reliable index for predicting the activity and severity in moderate and high disease activity of RA.

This single-center cross-sectional study included 139 inpatients diagnosed with RA according to 2010 European League Against Rheumatism (EULAR) criteria [17]. These patients were admitted into the Department of Rheumatology and Immunology at the Second Xiangya Hospital of Central South University from 01 July,2016 to 30 August,2019. Patients with other rheumatic diseases, severe infections, or malignancies were excluded from this study. Baseline demographic and clinical data were collected retrospectively from their medical records. At the same time, the levels of rheumatoid factors (RF-IgA, RF-IgG, and RF-IgM), levels of anti-cyclic citrullinated peptide antibody (CCP), blood index including mean platelet volume (MPV) and red blood cell distribution width (RDW) were measured, and PLR and NLR were calculated. The joint function was evaluated according to the standards established by the American Academy of Rheumatology [18].
The disease activity of RA was determined using DAS 28 system. To calculate DAS28-ESR and DAS28-CRP, the tender joint count (TJC), swollen joint count (SJC), erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP) levels were measured, and the patients' global assessment (PGA) in visual analog scale (VAS) were applied. Because patients with low disease activity were not admitted to the inpatient ward of the university hospital, the enrolled patients were assumed to have moderate and high disease activity,49 patients whose DAS28-CRP ranged from 3.2 to 5.1were categorized into the moderate disease activity group, and 90 patients whose DAS28-CRP was higher than 5.1were categorized into the high disease activity group. The study was approved by the Institutional Ethics Committee of the Second Xiangya Hospital of Central South University (2020-K011) and was conducted according to the Declaration of Helsinki. Written informed consents were obtained from all participants.

Ultrasound evaluation
To assess joint severity, joint ultrasound scores, including GS-US and PD-US, were measured using a Esote MyLab30Gold color Doppler ultrasound system. The ultrasound test was performed and veri ed for knee joints and/or wrist joints by the 2 ultrasound physicians who was unknown about patient's clinical data. Joints numbers of ultrasound tests ranged from 1 to 4 were decided by the patients' willingness to bear the medical costs. The total joint ultrasound score (T-US) was calculated by adding GS-US and PD-US for the examined joint and obtaining the average score.
The most frequently used semi-quantitative methods were applied for scoring synovitis [19]. The joint in ammation activity was graded using GS-US as follows: Grade 0, normal joint with no synovial thickening; Grade 1, minimal synovial thickening ( lling the angle between the periarticular bones, without bulging over the line linking tops of the bones); Grade 2, synovial thickening bulging over the line linking the tops of the periarticular bones but without extension along the bone diaphysis; Grade 3, synovial thickening bulging over the line linking tops of the periarticular bones and with extension to at least one of the bone diaphysis. Furthermore, arthritis severity was semiquantitatively graded using PD-US as follows: Grade 0, no ow in the synovium; Grade 1, single vessel signals; Grade 2, con uent vessel signals in less than half of the synovial area; Grade 3, vessel signals in more than half of the synovial area.

Statistical analyses
Statistical analyses were performed using SPSS software version 26.0 (Chicago, IL, USA). To test the normality of the data, the Kolmogorov-Smirnov test was used. The normally distributed variables were expressed as the mean ± standard deviation (x ± s). Signi cant differences between groups were assessed using the t-test. Non-normally distributed variables were represented by the median (upper and lower quartiles), and the Mann-Whitney U-test was used to compare differences between two groups. The data were expressed in frequencies or percentages (%), and the Chi-square test was used to compare differences between groups. The Pearson's correlation test was used to assess the correlation between routine hematological indices with DAS28 or joint ultrasound scores (PD-US, GS-US, and T-US). Then, Logistic regression analyses were performed to predict the relationship between US or DAS28 and blood indices and to identify the blood indices that best re ect DAS28 and US. Finally, receiver operating characteristics (ROC) analysis was performed to compare the e cacy of blood indices, ESR, or CRP in re ecting the disease activity and severity of RA. Values with two-sided p-values of less than 0.05 were considered statistically signi cant.

Clinical characteristics of patients in different disease activity groups
The mean age of the patients was 58 years (23-84 years), and majority patients were females (100/139, 71.9%). There were no signi cant differences in terms of sex ratio, age, disease duration, number of joints tested, positivity ratio of RF-IgG, IgA, IgM, and anti-CCP antibody between patients with moderate and high DAS28 (p>0.05). The values of clinical indices (SJC, TJC, Joint function, PGA, ESR, and CRP) and ultrasound indices (T-US and PD-US) in patients with high activity RA were higher compared to patients with moderate activity RA. Moreover, patients with high disease activity had signi cantly elevated levels of routine hematological indices, including lymphocyte (L) count, platelet (PLT) count, MPV, PLR, and NLR compared to patients with moderate disease activity (p<0.05) (    Abbreviations: CCP, anti-cyclic citrullinated peptide antibody; CRP, C-reactive protein; DAS28, disease activity score in 28 joints; ESR, erythrocyte sedimentation rate; GS-US, grey-scale ultrasound score; Hb, hemoglobin; L, lymphocyte; MPV, mean platelet volume; N, neutrophil; NLR, neutrophil-to-lymphocyte ratio; PD-US, power Doppler ultrasound score; PGA, patient global assessment; PLR, plateletto-lymphocyte ratio; Plt, platelet; RDW, red blood cell distribution width; RF, rheumatoid factor; SJC, swollen joint count; TJC, tender joint count; T-US, total ultrasound score; WBC, white blood cell.   Table 2). However, no signi cantly correlations were observed between WBC, N, RDW, and DAS28. / : Because there were no signi cant differences in Pearson's correlation analysis, Logistic analysis was not applied. N/A: Because strong multicollinearity with other independent variables were observed, L was not applied with Logistic analysis.
Binary logistic regression (LR) analysis was performed to assess which of above markers were risk factors associated with the DAS28 grade. Because there was collinearity between lymphocyte with other indicators (tolerance <0.2), lymphocyte was excluded outside of LR analysis, therefore the variables included in the LR analysis were Hb, PLT, MPV, NLR, and PLR. The results con rmed that PLR, which correlates with DAS28-ESR (odds ratio [OR]=1.009, p=0.033) and DAS28-CRP (OR=1.007, p=0.025), was an independent risk factor for the disease activity of RA (Table 2).

Correlation of routine hematological indices with ultrasound scores
The correlation between routine hematological indices (Hb, WBC, N, L, PLT, RDW, MPV, NLR, and PLR) and joint ultrasound was assessed by Pearson's correlation analysis. Results of Pearson's correlation analysis showed that PLT, NLR, and PLR were signi cantly positively correlated with PD-US meanwhile PLT and PLR were signi cantly positively correlated with T-US (p<0.05). However there was no hematological index was found to be signi cantly correlated with GS-US (Figure 2-C, Figure 2-D, Figure 2-E, Table 2).
Binary logistic regression (LR) analysis was performed to assess which of above markers was risk factors associated with the joint ultrasound grade. The variables included in the LR analysis of PD-US grade (≥1, <1) were PLT, NLR, and PLR, while PLT and PLR were included in the LR analysis of T-US grade(≥3, <3). All variables were numerical and were entered simultaneously. Results of collinearity diagnostics indicated no collinearity between the variables (tolerance > 0.2 and VIF <5). The results con rmed that PLR, which correlates with ultrasound scores including PD-US (OR=1.016, p=0.012) and T-US, (OR=1.016, p=0.006), was an independent risk factor for the disease severity of RA (  Figure 3A). Then, by analyzing the ROC curve of PLR with that of DAS28-CRP as a standard, we found that PLR could re ect disease activity (DAS28-CRP) e ciently (AUC=0.718, p<0.000).
The cut-off value of PLR was 247.3, with high sensitivity (0.61) and speci city (0.78). Comparison of the ROC values of PLR+ESR, PLR+CRP, and PLR+ESR+CRP with the ROC value of PLR alone revealed that the AUROC was not signi cantly improved (0.726,0.715,0.728 and 0.718, p>0.05) ( Figure 3B). Therefore, PLR alone is a reliable index with good diagnostic e cacy for predicting disease activity.
Similarly, by comparing the ROC curve of PLR with that of PD-US as a standard, we found that PLR could re ect RA severity (PD-US) e ciently (AUC=0.789, p<0.000). The cut-off value of PLR was 228.1, with high sensitivity (0.61) and speci city (0.87). Comparison of the ROC values of PLR+ESR, PLR+CRP, and PLR+ESR+CRP with the ROC value of PLR alone revealed that the AUROC was not signi cantly improved (0.808, 0.819, 0.823, and 0.789, p>0.05) ( Figure 3C). Then, by comparing the ROC curve of PLR with that of T-US as a standard, we found that PLR could re ect RA severity (T-US) e ciently (AUC=0.799, p<0.000). The cut-off value of PLR was 228.1, with high sensitivity (0.61) and speci city (0.86). Comparison of the ROC values of PLR+ESR, PLR+CRP, and PLR+ESR+CRP with the ROC value of PLR alone revealed that the AUROC was not signi cantly improved (0.807,0.827,0.824 and 0.799, p>0.05)( Figure 3D). Therefore, PLR alone is a reliable index with good diagnostic e cacy for predicting disease severity.
According to the optimal cut-off value of PLR (average = 236.6), the patients were divided into low PLR and high PLR groups. Observing the changes in the clinical parameters of the two different PLR groups, DAS28, T-US, and PD-US were signi cantly different between the two groups (p<0.001) ( Table 3). Table 3 Comparison of the characteristics of disease activity score in 28 joints (DAS28) and ultrasound score between the two groups based on the cut-off value of platelet-to-lymphocyte ratio (PLR). Abbreviations: CRP, C-reactive protein; DAS28, disease activity score in 28 joints; ESR, erythrocyte sedimentation rate; GS-US, greyscale ultrasound score; PLR, platelet-to-lymphocyte ratio; T-US, total ultrasound score.

Discussion
Currently, the DAS system and imaging analysis, including ultrasonography and radiology, are widely used to evaluate the activity and severity of RA. However, there is a lack of a reliable hematological index that can be used to assess RA activity and severity with a highly convenient and cost-effective manner during routine hematological tests which would bene t developing countries with poor health infrastructure and lack of access to sophisticated imaging techniques. Therefore, we aimed to explore valuable indices used in simple routine blood tests that may evaluate both the activity and severity of RA. In this single-center retrospective study of 139 patients, we found that PLR was signi cantly correlated with the disease activity and joint severity of patients with RA, similar to DAS28 and ultrasound scores. Analysis of ROC indicated that PLR had a higher sensitivity and speci city than ESR and CRP in predicting disease activity and severity. As an economical and convenient clinical parameter, PLR can be a simple and reliable index for predicting activity and severity in moderate and high disease activity of RA.
In recent years, some routine hematological indices have been proposed as systemic in ammatory indices in rheumatic diseases. Several studies have shown that patients with systemic lupus erythematosus (SLE) have signi cantly increased NLR and PLR, which were positively correlated with the SLE disease activity index (SLEDAI) and that NLR may be an indicator for monitoring disease activity and re ecting renal involvement in patients with SLE [20]. Some studies have shown that the Psoriasis Area and Severity Index scores of psoriasis patients were signi cantly higher in the NLR-high and PLR-high subgroups than that in the NLR-low and PLR-low subgroups, The NLR-high subgroup showed higher CRP levels [21]. There were several studies have shown that NLR and PLR were signi cantly correlated with DAS28 in patients with RA and that PLR may be used as a new biomarker for predicting and diagnosing RA and RA-ILD [22,23]. There were a lot of papers showing that platelet activation and high platelet reactivity induce chronic in ammation through the production of thrombopoietin and cytokines. Platelets have been reported to participate in immune regulation and cytokine production depending on their activation state [24][25][26][27][28][29][30]. PLT and PLR was positively correlated with DAS28 and PLR was found as an in ammatory factor that can predict disease activity in our study. Moreover, our study showed that the count of lymphocytes in peripheral blood was negatively correlated with disease activity. The decrease in the count of peripheral lymphocytes may be caused by the continuous accumulation of lymphocytes in in ammatory joints and the high levels of early apoptosis markers in lymphocytes, which may be related to autoimmunity [31,32]. Our ndings were consistent with the above mechanism of RA.
Histologically, RA is characterized by in ammatory cell in ltration and the formation of rich vascular pannus. Angiogenesis starts at a very early stage of RA under the in uence of various in ammatory cytokines, chemokines, and growth factors [33]. High levels of Ang-1, Ang-2, VEGF, and MMP-3 in the serum and synovium of patients with RA re ect disease activity [34,35]. Compared to composite disease activity indices, PD-US vascularity grading is better correlated with some markers of angiogenesis and in ammation [3]. In recent years, PD-US has been considered an important tool to detect early synovial in ammation and monitor treatment response [36,37]. One clinical report in 2020 had shown that NLR and PLR were signi cantly positively correlated with all US parameters of disease activity (gray-scale US, power Doppler US, and global scores) [39]. In our study, the results demonstrated that PLR was an independent risk factor of ultrasound scores (PD-US and T-US). Mechanism of pathological process could help us to understand the nding from ultrasound. In in amed joints, activated circulating platelets can interact with erythrocytes, neutrophils, and lymphocytes in the vessel lumen at sites of vascular damage [40,41]. Platelets also promote the re-aggregation of neutrophils into the RA synovial uid. Furthermore, platelets can interact with neutrophils via the expression of platelet-type lipoxygenase and the activation of the eicosanoid pathway, thus aggravating synovitis [28]. Additionally, platelets and their activated products can activate the coagulation and brinolysis system of the local synovium of RA, induce VEGF secretion in in ammatory sites, increase vascular permeability, promote neovascularization, increase the expression of MMP-3, destroy and decompose cartilage matrix, and aggravate the development of RA [26,27]. All these studies provided strong evidence on the correlation between PLR and severity of RA (PD-US).
In clinical application, different scores based on different numbers of joints subjected to ultrasonic detection have been proposed, such as 28-joint, 12-joint, and 7-joint ultrasound scores [42][43][44], to effectively evaluate the severity of RA. However, examination of a large number of joints is associated with a longer detection time and higher medical costs, which greatly reduce the convenience of ultrasonography as a disease activity assessment tool in the real world. Several studies have shown that detecting a smaller number of joints is equally effective in assessing the disease activity and severity of RA [45]. In this study, we selected either the wrist or knee joint of patients with RA and identi ed the tested site guided by the patients' willingness, which was in line with the core principle of a patientreported outcome (PRO). Based on this basic principle, only the most affected joints can be selected and subjected to ultrasonography instead of examining pre-arranged numbers of joints. Therefore, patients can avoid undergoing an ultrasonic examination of a large number of joints, thus effectively reducing the time and medical costs associated with ultrasonography.
There were some limitations to this study. First, this was a cross-sectional study with a small sample size. Second, all enrolled patients come from hospitalized ones with moderate or high disease activity and there were no low disease activity patients which limited. We could not directly conclude that PLR was marker for the activity and severity of RA. Finally, longitudinal studies should be conducted to evaluate the clinical value of PLR in the prediction of treatment effects and relapse of RA in the future.

Conclusion
In this study, we mainly studied the correlations between routine hematological indices with the disease activity score and joint ultrasound score in patients with RA. We attempted to study if routine hematological indices could accurately assess the activity and joint severity of RA in a cost-effective and convenient manner. The results showed that PLR, NLR, PD-US, and T-US were signi cantly different between moderate and high disease activity groups. PLR, a clinical index routinely measured in hematological tests, was signi cantly correlated with PD-US and DAS28. Moreover, the results of logistic regression analysis showed that PLR is an independent risk factor for disease activity and joint severity in RA. The e cacy of using PLR alone to evaluate disease activity and severity was similar to that of using combined CRP and ESR. PLR is a blood parameter that can be economically and conveniently assessed during routine blood tests in developing countries, PLR can be a simple and reliable index for predicting the activity and severity in moderate and high disease activity of RA.

Data availability statement
Additional data are available upon reasonable request. The datasets used and/or analyzed in the present study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
The study was approved by the Institutional Ethics Committee of the Second Xiangya Hospital of Central South University (2020-K011) and was conducted according to the Declaration of Helsinki. Written informed consents were obtained from all participants.

Consent for publication
Not applicable.