Immune dysregulation driven by elevated platelet-to-lymphocyte ratio aggravates myasthenia gravis

Abstract

Objective

Previous studies have suggested a potential association between the platelet-to-lymphocyte ratio and disease activity in myasthenia gravis. However, the immunological mechanisms underlying this association remain insufficiently elucidated.

Methods

A retrospective cohort of 229 patients with myasthenia gravis and a single-cell RNA sequencing dataset were analyzed to investigate the relationship between platelet-to-lymphocyte ratio and disease severity. Clinical associations were assessed using the Myasthenia Gravis Foundation of America classification and multivariable logistic regression, while single-cell RNA sequencing data were integrated to characterize immune alterations associated with elevated platelet-to-lymphocyte ratio.

Results

Patients with severe myasthenia gravis had longer disease duration and higher frequencies of bulbar symptoms, thymoma, and repetitive nerve stimulation positivity (all p < 0.001). Although median platelet-to-lymphocyte ratio values did not demonstrate significant groupwise differences (p = 0.108), multivariate analysis confirmed that an elevated platelet-to-lymphocyte ratio was independently associated with greater myasthenia gravis severity (adjusted odds ratio = 1.027, 95% confidence interval: 1.003–1.052, p = 0.034). Single-cell RNA sequencing revealed immune dysregulation in patients with a high platelet-to-lymphocyte ratio, characterized by increased platelets and neutrophils, reduced natural killer cells, and upregulation of platelet activation, cell–cell adhesion, and integrin-mediated signaling pathways, indicating a shift toward innate immune activation and impaired immune coordination.

Conclusion

Elevated platelet-to-lymphocyte ratio independently predicts myasthenia gravis severity and may reflect immune dysregulation that contributes to disease progression and neuromuscular junction dysfunction.

Keywords

Myasthenia gravis platelet-to-lymphocyte ratio single-cell RNA sequencing immune dysregulation retrospective cohort study

Introduction

Myasthenia gravis (MG) is an autoimmune disorder characterized by autoantibodies targeting postsynaptic components, leading to impaired neuromuscular junction (NMJ) transmission.¹ Clinically, MG presents with fluctuating skeletal muscle weakness and fatigability that typically worsen with exertion.² Approximately 80% of patients harbor detectable autoantibodies against the acetylcholine receptor (AChR), while smaller subsets exhibit antibodies directed against muscle-specific tyrosine kinase or low-density lipoprotein receptor-related protein 4.³ Although current therapeutic strategies for MG primarily aim to suppress inflammation and neutralize autoantibodies, clinical responses vary widely, underscoring the need for simple and reliable inflammatory markers to guide individualized patient care.⁴ The platelet-to-lymphocyte ratio (PLR), an easily accessible and cost-effective marker derived from routine complete blood counts, reflects the dynamic balance between two essential immune components: platelets, which initiate and propagate inflammatory responses, and lymphocytes, the principal mediators of adaptive immunity. Platelets are increasingly recognized as active immune effectors that contribute to inflammation, immune cell activation, and tissue injury in diseases such as systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), and multiple sclerosis (MS).⁵ Previous studies have suggested an association between PLR and MG; however, the immunological mechanisms underlying this relationship remain poorly defined.^6–9

Advances in single-cell RNA sequencing (scRNA-seq) enable high-resolution profiling of peripheral immune cells, revealing detailed characterization of activation states, lineage relationships, and signaling networks. Whether elevated PLR in MG reflects dysregulation of specific immune signaling pathways remains unclear.^10,11 To the best of our knowledge, no studies have systematically integrated scRNA-seq data with clinical cohorts to investigate the immunological basis of PLR elevation in MG, particularly with respect to altered intercellular immune communication.

Accordingly, we evaluated the association between PLR and MG severity and explored its immunological context through the integration of clinical and molecular evidence. We conducted a retrospective analysis of a clinical cohort to examine whether PLR is associated with the Myasthenia Gravis Foundation of America (MGFA)-defined severity, and we used scRNA-seq to characterize the immune landscape associated with elevated PLR, with particular emphasis on intracellular communication. These findings may provide hypothesis-generating insights into immune-based strategies and identify potential therapeutic targets in MG.

Methods

To systematically investigate the clinical utility of PLR in assessing disease severity in MG and to further explore its potential mechanistic underpinnings, this study was designed from two complementary perspectives: (a) a retrospective cohort study based on clinical data to compare disease severity across different PLR levels, and (b) a scRNA-seq analysis to compare patients with elevated versus low PLR with respect to immune cell composition, functional status, and molecular pathway activation. The clinical cohort (229 patients) and the scRNA-seq dataset (10 patients; Gene Expression Omnibus (GEO) accession GSE227835) were analyzed independently, with the single-cell analysis intended to generate mechanistic hypotheses rather than to provide matched patient-level data. Through this integrative approach, we aimed to elucidate the immunological features associated with elevated PLR and provide mechanistic insights into its relevance to MG disease severity.

Study population and data collection

This retrospective study included 287 patients diagnosed with MG at the First Affiliated Hospital of Nanchang University between January 2018 and December 2024. For patients with multiple hospitalizations during the study period, only data from the first admission were analyzed. Inclusion criteria were as follows: (a) typical clinical manifestations consistent with MG; (b) at least one positive result on neostigmine testing, repetitive nerve stimulation (RNS), or AChR antibody testing; and (c) exclusion of alternative diagnoses prior to the confirmation of MG. Exclusion criteria were as follows: (a) incomplete clinical data; (b) severe renal or hepatic dysfunction; (c) malignancies other than thymoma; and (d) other autoimmune diseases. After application of these criteria, 229 patients were included in the final analysis (Figure 1).

Figure 1.

Flowchart demonstrating the selection of participants for the study (n = number of patients).

Clinical data were retrospectively collected, including demographic information, disease duration, RNS results, presence of thymoma, history of hypertension and diabetes, clinical manifestations of MG, and MGFA classification.⁷ In addition, all blood samples, including those for complete blood count, serum albumin, and C-reactive protein (CRP), were collected in the morning after an overnight fast to ensure consistency across laboratory results. MGFA classification was recorded as a measure of disease severity and was documented at the time of hospital admission, corresponding to the peak of clinical severity. Based on the MGFA classification, patients were categorized into two groups: mild (MGFA types I and II) and severe (MGFA types III–V). The mild group comprised patients with ocular or mild generalized muscle weakness, while the severe group comprised patients with moderate-to-severe weakness, including those requiring mechanical ventilation.⁹ Quantitative MG (QMG) scores were not systematically available in this retrospective dataset and were therefore not included in the analysis; this limitation is acknowledged in the Discussion section. This study was approved by the Institutional Ethics Committee of the First Affiliated Hospital of Nanchang University (Approval No. AF-SG-03-2.1-IIT). As this was a retrospective study using deidentified data, informed consent from individual patients was not required. All patient information was anonymized to ensure confidentiality, and no data were shared with third parties. The study was conducted in accordance with the Declaration of Helsinki (1975, revised 2024).

Statistical analysis

All statistical analyses were performed using SPSS software (version 26.0). The normality of continuous variables was assessed using the Shapiro–Wilk test. For non-normally distributed continuous variables, the Mann–Whitney U test was used for between-group comparisons. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. PLR was treated as a continuous variable in all regression analyses. Univariate and multivariate logistic regression analyses were then performed to investigate the association between PLR and MG severity, with adjustment for confounders. Variables showing statistical significance in univariate analysis, along with clinically relevant factors, were included in the multivariate model. All statistical tests were two-sided, and a p value <0.05 was considered statistically significant.

Single-cell RNA sequencing analysis

Single-cell transcriptomic datasets were obtained from the GEO database (http://www.ncbi.nlm.nih.gov/geo/). Specifically, samples from 10 AChR antibody-positive patients with MG were obtained from the dataset GSE227835. After quality control, a total of 33,660 cells were retained. The scRNA-seq data were processed using the Seurat R package (version 5.0.2).⁸ Quality control procedures retained cells meeting the following criteria: (a) number of detected genes per cell (nFeature_RNA) between 800 and 2500, (b) total transcript counts (nCount_RNA) between 2500 and 10,000, and (c) mitochondrial gene content (percent.mt) <7%. For principal component analysis, the FindVariableFeatures function in Seurat was used to identify the top 2000 highly variable genes. Cell clustering was performed using the FindNeighbors and FindClusters functions, and nonlinear dimensionality reduction was conducted using the RunUMAP function with default parameters. Marker genes for each cluster were identified using the FindAllMarkers function (parameters: logfc.threshold = 0.25, only.pos = TRUE, and min.pct = 0.25). Cell types were manually annotated based on established markers reported in the literature and the CellMarker database.⁹

In the clinical dataset, PLR was defined as the ratio of the absolute platelet count to the absolute lymphocyte count. In the absence of absolute cell counts in the scRNA-seq data, PLR was approximated as the ratio of the relative abundance of platelet-related cells to lymphocytes among peripheral blood mononuclear cells (PBMCs), serving as a surrogate measure of systemic platelet–lymphocyte balance. The PLR approximation was calculated using the following formula:

$PLR= \frac{\begin{matrix} Relative abundance of \\ platelet - related cells \end{matrix}}{\begin{matrix} Relative abundance of \\ lymphocyte cells \end{matrix}}$

For each patient sample, PLR was calculated based on its immune cell composition. The mean PLR across all samples was used as the threshold to stratify patients into “high PLR” and “low PLR” groups, rather than ROC-derived cutoffs. Samples with PLR values above the overall mean were categorized as the “High-PLR” group (n = 4 samples, 11,288 cells), whereas those below the mean were categorized as the “Low-PLR” group (n = 6 samples, 22,372 cells). Group-level cell-type compositions are summarized in Table S1. To compare the distribution of immune cell subsets between groups, two-way stratification was performed based on “PLR group” and “cell type.” The frequency of each immune cell type was calculated within each group and expressed as a proportion relative to the total number of cells in that group. Differential gene expression analysis was performed using Seurat’s FindMarkers function across all cells. Significantly differentially expressed genes (DEGs) were filtered using the criteria logfc.threshold >0.25 and min.pct > 0.25.

To explore the biological functions of the DEGs, gene ontology (GO) enrichment analysis was performed using the clusterProfiler package. Only upregulated DEGs were included in the analysis, and gene symbols were converted to Entrez IDs. The organism database was set to org.Hs.eg.db (humans), and enrichment was restricted to the biological process category. Multiple testing correction was applied using the Benjamini–Hochberg method, with statistical significance thresholds defined as adjusted p value (p.adjust) < 0.05 and a q-value < 0.05.

To investigate alterations in intercellular immune communication between distinct PLR groups, we performed cell–cell communication analysis using the CellChat R package (version 1.6.1) on the integrated single-cell dataset. First, cell subpopulation information from the integrated Seurat object (merged_seurat) was assigned to cell-type labels, and the metadata were verified to include both group (PLR stratification) and cell-type annotations. Based on the study species (humans), the corresponding CellChatDB and protein–protein interaction databases were used to identify ligand–receptor pairs and associated signaling pathways. A reference gene set comprising all ligands and receptors curated in the database was constructed to further filter the expression matrix for analysis.

For each group (“High PLR” and “Low PLR”), the corresponding expression data and metadata were extracted, and CellChat objects were constructed according to the following workflow: (a) extraction of the expression matrix and initialization of the createCellChat object; (b) identification of overexpressed genes and ligand–receptor interaction pairs; (c) projection of pathways by integrating the protein–protein interaction network; (d) computation of cell–cell communication probabilities using the computeCommunProb function; (e) filtration of unreliable communication pairs based on a minimum threshold (min.cells = 10); (f) identification of enriched signaling pathways using computeCommunProbPathway, followed by network aggregation with aggregateNet; and (g) exportation of ligand–receptor communication data and signaling pathway networks for each group separately.

Subsequently, CellChat objects from the “High PLR” and “Low PLR” groups were merged using the mergeCellChat function to compare the number (count) and strength (weight) of communications between groups. The rankNet function was applied to generate bar plots for visual comparison. To further validate expression differences in key ligand–receptor pairs within representative signaling pathways, several representative genes (e.g. MIF, CD45, adhesion pathways, and T/B cell activation pathways) were selected. DotPlots were generated to display their expression patterns across different cell subpopulations, with results compared between groups.

Results

Characteristics of enrolled patients

A total of 229 patients with MG were included in this retrospective study. Among these patients, 122 had ocular MG, and 107 had generalized MG. According to the MGFA classification, 156 patients were classified as MGFA types I and II, and 73 as MGFA types III and V. The mean age was 49.18 ± 18.77 years in the MGFA I and II group and 50.15 ± 15.71 years in the MGFA III–V group. There were 75 male patients (48.08%) in the MGFA I and II group and 30 (41.10%) in the MGFA III–V group. Patients in the MGFA III–V group had a significantly longer disease duration, with a median duration of 12.00 (3.00, 39.00) months, compared to 2.00 (0.78, 8.00) months in the MGFA I and II group. Significant differences were also observed between the two groups in RNS positivity, the presence of bulbar symptoms, and thymoma (all p < 0.001). For routine blood tests, lymphocyte counts differed significantly between the two groups (p = 0.01). Moreover, both the neutrophil-to-lymphocyte ratio (NLR) and lymphocyte-to-monocyte ratio (LMR) differed significantly between groups (p = 0.037 and p = 0.023, respectively).

Median PLR values demonstrated numerical differences between groups stratified by disease severity. Specifically, the median PLR was 117.84 (101.94, 168.17) in the severe PLR group and 119.72 (85.85, 147.60) in the control group. However, this difference did not reach statistically significance (p = 0.108), suggesting that PLR distribution alone may be insufficient to establish a clear association with disease severity. No significant differences were observed between the two groups with respect to age, age at onset, sex, comorbid hypertension or diabetes, infection status, monocyte count, platelet count, neutrophil count, albumin, or CRP (Table 1).

Table 1.

Comparisons of baseline characteristics between MGFA III–V groups and MGFA I–II groups.

Variable	MGFA I–II (n = 156)	MGFA III–V (n = 73)	p value
Age (years), mean ± SD	49.18 ± 18.77	50.15 ± 15.71	0.222
Age of onset (years), mean ± SD	47.22 ± 20.10	46.85 ± 18.34	0.266
Disease duration (months), median (IQR)	2.00 (0.78, 8.00)	12.00 (3.00,39.00)	0.021*
Male, n (%)	75 (48.08%)	30 (41.10%)	0.323
Bulbar symptoms, n (%)	16 (10.26%)	48 (65.75%)	<0.001*
RNS, n (%)	59 (37.82%)	38 (52.06%)	<0.001*
Hypertension, n (%)	15 (9.62%)	4 (5.48%)	0.29
Diabetes, n (%)	4 (2.56%)	2 (2.74%)	0.938
Thymoma, n (%)	23 (14.74%)	26 (35.62%)	<0.001*
Infection	12 (7.69%)	6 (8.22%)	0.876
WBCs, median (IQR)	6.30 (5.07, 7.48)	5.59 (4.41, 7.15)	0.545
Platelets, median (IQR)	218.50 (185.00, 248.00)	223.50 (199.50, 241.75)	0.596
Lymphocytes, median (IQR)	1.84 (1.50, 2.42)	1.60 (1.27, 2.17)	0.01*
Monocytes, median (IQR)	0.43 (0.33, 0.52)	0.36 (0.29, 0.53)	0.909
Neutrophils, median (IQR)	3.73 (2.87, 4,38)	3.37 (2.55, 5.02)	0.912
Albumin, median (IQR)	41.85 (39.33,44.60)	41.65 (39.55, 43.93)	0.05
CRP, median (IQR)	1.33 (0.60, 1.95)	1.08 (0.49, 2.33)	0.698
NLR, median (IQR)	2.03 (1.47, 2.45)	1.93 (1.54, 2.93)	0.037*
PLR, median (IQR)	119.72 (85.85,147.60)	117.84 (101.94, 168.17)	0.108
LMR, median (IQR)	4.87 (3.88, 6.24)	4.64 (3.29, 5.70)	0.023*

CRP, C-reactive protein; LMR, lymphocyte-to-monocyte ratio; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; RNS, repetitive nerve stimulation; WBC, white blood cell counts.

Association between MG severity and PLR

Univariate logistic regression analysis demonstrated that the presence of bulbar symptoms (odds ratio (OR) = 16.8, 95% confidence interval (CI): 8.276–34.102, p < 0.001), positive RNS findings (OR = 3.663, 95% CI: 1.875–7.157, p < 0.001), and the presence of thymoma (OR = 3.391, 95% CI: 1.755–6.553, p < 0.001) were significantly associated with greater MG severity. Among laboratory parameters, serum albumin (OR = 0.914, 95% CI: 0.843–0.991, p = 0.029), PLR (OR = 1.007, 95% CI: 1.002–1.012, p = 0.004), and NLR (OR = 1.352, 95% CI: 1.088–1.681, p = 0.006) were also significantly associated with MG severity. In contrast, no significant associations were observed for sex, age, age at onset, disease duration, hypertension, diabetes, infection, white blood cell count, platelet count, lymphocyte count, monocyte count, neutrophil count, CRP, or LMR (Table 2).

Table 2.

Univariate logistic regression analyses of factors for severe disease of MG.

	OR	95% CI	p value
Age	1.013	0.996–1.03	0.126
Gender	0.753	0.429–1.322	0.324
Age of onset	1.011	0.996–1.026	0.164
Disease duration	1.002	0.998–1.006	0.405
Bulbar symptoms	16.8	8.276–34.102	<0.001*
RNS	3.663	1.875–7.157	<0.001*
Thymoma	3.391	1.755–6.553	<0.001*
Infection	1.091	0.392–3.032	0.868
WBCs	1.026	0.899–1.171	0.706
Platelets	0.999	0.994–1.004	0.679
Lymphocytes	0.716	0.479–1.071	0.104
Monocytes	1.128	0.206–6.181	0.889
Neutrophils	1.118	0.959–1.305	0.154
Albumin	0.914	0.843–0.991	0.029*
CRP	1.077	0.984–1.180	0.108
NLR	1.352	1.088–1.681	0.006*
PLR	1.007	1.002–1.012	0.004*
LMR	0.921	0.806–1.053	0.227

CI: confidence interval; CRP: C-reactive protein; LMR: lymphocyte-to-monocyte ratio; MG: myasthenia gravis; NLR: neutrophil-to-lymphocyte ratio; OR: odds ratio; PLR: platelet-to-lymphocyte ratio; RNS: repetitive nerve stimulation; WBCs: white blood cell counts.

To adjust for potential confounders, multivariate logistic regression analyses were performed, incorporating both statistically significant variables from univariate analyses and clinically relevant covariates (Table 3). Three separate models were constructed. In Model 1, which included PLR, sex, age, age at onset, and disease duration, PLR remained significantly associated with MG severity (OR = 1.008, 95% CI: 1.003–1.013, p = 0.003). This association persisted in Model 2, which additionally adjusted for thymoma and RNS results (OR = 1.012, 95% CI: 1.004–1.019, p = 0.002). Notably, Model 3—the fully adjusted model including PLR, sex, age, age at onset, disease duration, thymoma, RNS, platelet count, lymphocyte count, neutrophil count, NLR, and comorbidities (hypertension and diabetes), also confirmed a significant association between PLR and MG severity (OR = 1.027, 95% CI: 1.003–1.052, p = 0.034). Collectively, these findings indicate that PLR is independently and significantly associated with disease severity in patients with MG.

Table 3.

Associations between PLR and disease severity of MG.

PLR	OR	95% CI	p value
Adjusted Model 1	1.008	1.003–1.013	0.003*
Adjusted Model 2	1.012	1.004–1.019	0.002*
Adjusted Model 3	1.027	1.003–1.052	0.034*

CI: confidence interval; MG: myasthenia gravis; OR: odds ratio; PLR: platelet-to-lymphocyte ratio.

Immune mechanism of MG progression associated with elevated PLR

Single-cell transcriptomic analysis identified eight major peripheral immune cell populations: T cells, B cells, monocytes, neutrophils, natural killer (NK) cells, NK T cells, platelets, and erythrocytes. Uniform Manifold Approximation and Projection (UMAP) visualization delineated the distribution of these immune subsets, demonstrating well-defined and clustering across samples (Figure 2(a)). Stratification by PLR levels revealed significant alterations in immune cell composition in the high-PLR group (Figure 2(b)). Specifically, the relative abundance of platelets was higher in the high-PLR group (1.6%) compared with the low-PLR group (0.3%). In contrast, NK cells and monocytes were reduced (4.2% vs. 9.0% and 0.04% vs. 0.14%, respectively), whereas T cells were slightly increased in the high-PLR group (71.3% vs. 64.6%). These findings suggest that elevated PLR is associated with a reshaped peripheral immune landscape, particularly characterized primarily by increased platelet abundance and a reduced NK cell proportion.

Figure 2.

Single-cell RNA sequencing reveals altered immune cell composition and functional states in elevated PLR patients with MG. (a) UMAP plot visualizing the identified major peripheral blood immune cell types in patients with MG. (b) Bar chart showing the relative proportion of major immune cell types in peripheral blood, comparing patients with elevated PLR and low PLR and (c) GO enrichment analysis of DEGs in the PLR elevated group. Bar plots display the top significantly enriched biological processes, the size and color of dots indicate the gene count and adjusted p value, respectively. DEG: differentially expressed gene; GO: gene ontology; MG: myasthenia gravis; PLR: platelet-to-lymphocyte ratio; UMAP: Uniform Manifold Approximation and Projection.

To investigate the biological processes associated with elevated PLR, GO enrichment analysis was performed on DEGs identified in the high-PLR group (Figure 2(c)). The enriched terms included multiple immune-related processes, most notably “homotypic cell–cell adhesion” and “platelet aggregation,” both of which were highly significant (adjusted p < 0.01). Additional enriched pathways included “wound healing,” “regulation of cell–cell adhesion,” “integrin-mediated signaling,” “platelet activation,” and “regulation of endothelial cell migration.” These results suggest that elevated PLR may contribute to MG progression by enhancing platelet function and promoting immune cell adhesion, thereby modulating the inflammatory microenvironment.

To further elucidate PLR-associated intercellular communication patterns, CellChat was used to compare global signaling networks between high- and low-PLR groups (Figure 3(a) and (b)). The low-PLR group exhibited stronger overall information flow and a greater number of interactions across several signaling pathways, including MIF, MHC-I, APP, CD99, and ICAM. These findings indicate a more extensive and coordinated immune communication network in the low-PLR group. Conversely, the high-PLR group displayed a simplified communication network, with enhanced signaling limited to a number of pathways, such as CD39 and GALECTIN, reflecting an overall impairment in immune coordination.

Figure 3.

Altered immune cell communication networks and key molecular expression in elevated PLR patients with MG. (a) Bar plot illustrating the overall information flow within the immune cell communication network. (b) Bar plot showing the comparison of information flow strength for specific key signaling pathways between PLR elevated and PLR reduced groups and (c) DotPlot displaying the expression levels of key ligand–receptor molecules across different immune cell subpopulations, stratified by PLR group. MG: myasthenia gravis; PLR: platelet-to-lymphocyte ratio.

To validate the molecular basis underlying these intercellular communications, the expression levels of key ligand–receptor pairs were analyzed in both groups (Figure 3(c)). In the low-PLR group, multiple molecules involved in adaptive immunity (CD6, CXCR4, PTPRC, TNFRSF13C, and CD22) were more highly expressed in T cells, B cells, and NK cells, suggesting enhanced immune activation and effector function. In contrast, in the high-PLR group, these adaptive immune-related molecules were generally downregulated, while proinflammatory factors such as MIF, ICAM1, and SELPLG were significantly upregulated in monocytes and platelets, indicating aberrant activation of innate immune pathways.

Based on these findings, we propose an immunopathogenic model (Figure 4). In this model, elevated PLR reflects immune imbalance, characterized by enhanced abundance and activation of innate immune cells (platelets and neutrophils) together with impaired function of adaptive immune cells, despite an increased proportion of T cells. In the high-PLR group, platelets exhibited elevated expression of adhesion molecules (e.g. SELPLG and ICAM1), suggesting enhanced capacity for immune cell interactions and vascular inflammation. Although reduced in proportion, monocytes retained high expression of proinflammatory cytokines such as MIF, indicating sustained innate immune activation. Meanwhile, T and B cells exhibited downregulation of key functional markers (e.g. CD6, CXCR4, and TNFRSF13C), suggesting functional exhaustion or impaired immune regulation. Collectively, these alterations reflect a state of immune dysregulation, characterized by innate overactivation, impaired adaptive coordination, and disrupted intercellular communication, which may facilitate persistent autoantibody production and promote NMJ damage.

Figure 4.

Proposed immune mechanism model of MG progression driven by elevated PLR. The model depicts how elevated PLR leads to altered immune cell composition (increased platelets and neutrophils; decreased monocytes and NK cells; and slightly increased but dysfunctional T cells), enhanced innate immunity, impaired adaptive immunity, abnormal cell–cell interactions, and an imbalanced immune network, ultimately resulting in persistent autoantibody production, exacerbated neuromuscular junction damage, and MG disease progression. MG: myasthenia gravis; NK: natural killer; PLR: platelet-to-lymphocyte ratio.

Discussion

This study integrated retrospective clinical data from 229 patients with MG with peripheral blood single-cell transcriptomic data from 10 patients with AChR antibody-positive MG. We first demonstrated a significant association between elevated platelet counts, increased PLR levels, and greater disease severity. Subsequently, comprehensive scRNA-seq analysis was used to systematically evaluate the immunological relevance of PLR in MG. For the first time, we propose that elevated PLR may contribute to MG progression by altering immune cell composition and function, thereby disrupting intercellular communication networks and reshaping the activation patterns of key signaling pathways. These findings extend the clinical relevance of PLR by elucidating its potential mechanistic contributions to MG and providing insights into immune dysregulation in MG pathogenesis.

Previous studies have extensively explored the association between PLR and various neuroimmunological diseases, proposing PLR as a potential biomarker for disease severity and prognosis in inflammatory conditions.^6,7 Although emerging evidence suggests a possible link between PLR and disease activity in MG, the underlying immunological mechanisms remain insufficiently characterized.^8,9 In this study, we retrospectively analyzed clinical data from 229 patients with MG, stratified into mild (MGFA I and II) and severe (MGFA III and V) groups according to MGFA classification. At baseline, median PLR values did not differ significantly between the two groups. However, logistic regression analysis revealed a more pronounced association between PLR and disease severity. In the unadjusted univariate model, elevated PLR was associated with a significantly increased risk of severe MG. This association remained robust in the fully adjusted multivariate model (Model 3) after adjustment for potential confounders, including sex, age, disease duration, presence of thymoma, RNS positivity, platelet count, lymphocyte count, neutrophil count, NLR, and comorbidities such as hypertension and diabetes. Although the OR (1.007–1.027) appear numerically modest, they are statistically robust and may become clinically meaningful when PLR is considered a continuous inflammatory biomarker at a large population level. This observation is consistent with other inflammatory indices, in which even modest per-unit changes can translate into clinically meaningful shifts in risk at the population level. Moreover, although groupwise comparisons of median PLR did not reach statistical significance, its predictive value became evident when PLR was analyzed as a continuous variable in regression models. This apparent discrepancy likely reflects methodological differences between unadjusted groupwise tests and covariate-adjusted regression models. The Mann–Whitney U test assesses median differences without accounting for confounding factors, whereas multivariate regression improves sensitivity by adjusting for variables such as thymoma, RNS positivity, and NLR, thereby revealing the independent contribution of PLR to MG severity.^8,12

Beyond their classical role in hemostasis, platelets are now recognized as key inflammatory effectors that influence both innate and adaptive immune responses. Activated platelets possess thromboinflammatory properties, bridging coagulation and immune activation under various physiological and pathological conditions.^13,14 For example, in SLE, platelet hyperactivation promotes IFN-α production and exacerbates immune complex–mediated tissue injury¹⁵; in RA, platelet-derived microparticles activate synovial cells and contribute to chronic synovial inflammation¹⁶; and in neuroimmune disorders such as MS, platelets facilitate T cell migration into the central nervous system and promote lesion formation.¹⁷ Collectively, these findings highlight the immunomodulatory role of platelets in the pathogenesis and progression of autoimmune diseases. However, the specific contribution of PLR to the pathophysiology of MG remains poorly defined.

To further explore the immunological mechanisms underlying elevated PLR, we integrated scRNA-seq data to assess peripheral immune cell heterogeneity in patients with high versus low PLR values. Our data revealed an increased proportion of neutrophils and platelets in the high-PLR group. Functional enrichment analysis confirmed the upregulation of pathways related to platelet activation and cell adhesion, reflecting enhanced innate immune activity in the context of elevated PLR in patients with MG. In particular, enrichment of platelet activation and integrin-mediated signaling pathways indicates enhanced platelet–immune cell adhesion and vascular inflammation, which are recognized mechanisms linking platelet hyperactivity to autoimmune responses. Notably, platelets exhibited increased expression of adhesion molecules such as SELPLG and ICAM1, suggesting potential involvement in aberrant immune cell interactions and amplification of inflammatory responses.¹⁸ Although the proportion of monocytes was reduced in the high-PLR group, expression of proinflammatory mediators, such as MIF, remained elevated, indicating that monocytes may retain the capacity to promote inflammation and contribute to tissue damage.¹⁹ This innate immune-driven inflammatory milieu may create conditions that favor B cell activation and autoantibody production through cytokine-mediated signaling.²⁰

Furthermore, our results support the presence of impaired adaptive immune function and disrupted immune tolerance in patients with elevated PLR. Despite an increased proportion of T cells in the high-PLR group, expression of key functional molecules such as CD6 and CXCR4 was downregulated, suggesting reduced activation potential and effector capacity.^19,21 This functional impairment may reflect chronic inflammation–induced immune exhaustion or dysfunction of regulatory T cell subsets. Such deficits in adaptive immune function may directly impair the immune system’s ability to maintain tolerance and effectively modulate inflammatory responses.²⁰ Under normal conditions, finely tuned regulation of T and B cell responses—including regulatory T cell activity—is essential for preventing autoimmunity. Disruption of these mechanisms may compromise immune tolerance and permit uncontrolled inflammation, ultimately sustaining autoantibody production and promoting NMJ damage.²² CellChat analysis further demonstrated a simplified intercellular communication network in the high-PLR group. This reduction in signaling efficiency may reflect diminished immune coordination, thereby limiting precise regulation of immune responses. Collectively, these observations support the hypothesis that innate immunity overactivation combined with impaired adaptive immunity is associated with elevated PLR and may contribute to MG severity.

This is the first study to integrate retrospective clinical data with single-cell transcriptomic analysis in MG, demonstrating that elevated PLR is associated with distinct immune cell alterations and potential pathogenic mechanisms. These findings suggest that PLR may serve not only as a biomarker of disease severity but also highlight potential therapeutic targets, such as platelet-mediated inflammation and impaired adaptive immunity. The retrospective design may introduce confounding and selection bias; however, these effects were minimized through multivariate logistic regression incorporating clinically relevant covariates. MGFA classifications were assessed at hospital admission, corresponding to peak clinical severity. However, QMG scores were not systematically recorded, limiting the granularity of severity assessment. The cross-sectional design precludes inference regarding causal directionality. Although elevated PLR may promote immune dysregulation that aggravates MG, it is also plausible that disease severity itself elevates PLR through systemic inflammation. Prospective studies are required to clarify this bidirectional relationship. Several additional limitations warrant consideration. First, technical biases—such as differential cell capture efficiency, platelet fragility during PBMC isolation, and differences in RNA content—may have distorted relative abundances. In addition, platelets are typically underrepresented or lost during density gradient centrifugation, potentially leading to underestimation of their true contribution. This limitation underscores the need for orthogonal validation using flow cytometry. Second, although PLR may be influenced by non–disease-specific factors such as infection or pharmacological treatment, these confounders were minimized through predefined exclusion criteria and covariate adjustment. Notably, PLR remained significant in multivariate models, suggesting the presence of a potentially disease-specific immune mechanism. Future prospective studies are needed to validate these findings. Importantly, the clinical cohort and scRNA-seq dataset were independent, limiting direct translational inference. Nevertheless, this approach enabled the exploration of potential immune mechanisms underlying elevated PLR in MG. Given these limitations, our mechanistic observations should be regarded as hypothesis-generating and require validation in future experimental and longitudinal studies.

Conclusion

Elevated PLR is closely associated with disease severity in MG and serves as an independent predictive factor. Single-cell analysis reveals that increased PLR may reflect underlying immune dysregulation in patients with MG, potentially contributing to impaired NMJ function and disease progression. Collectively, these findings provide mechanistic insights that may inform future diagnostic and therapeutic strategies in MG.

Supplemental Material

sj-pdf-1-imr-10.1177_03000605251412108 - Supplemental material for Immune dysregulation driven by elevated platelet-to-lymphocyte ratio aggravates myasthenia gravis

Supplemental material, sj-pdf-1-imr-10.1177_03000605251412108 for Immune dysregulation driven by elevated platelet-to-lymphocyte ratio aggravates myasthenia gravis by Si Luo, Ziwei Song, Alina Zhawatibai, Yusen Qiu, Menghua Li, Yu Zhu, Yanyan Yu, Meihong Zhou and Daojun Hong in Journal of International Medical Research

Footnotes

Acknowledgments

We would like to acknowledge the authors of the publicly available dataset GSE227835 in the Gene Expression Omnibus database for making their single-cell RNA sequencing data accessible to the research community. Their contribution provided a valuable resource for our analysis of immune cell heterogeneity in myasthenia gravis.

Author contributions

SL, ZS, and AZ were responsible for acquisition of data, SL, YQ, YY, and YZ were responsible for drafting the manuscript. MZ and ML were responsible for study concept and design, MZ and DH were responsible for revising the manuscript. All authors have read and approved the final manuscript.

Data availability

The single-cell RNA sequencing data analyzed in this study were obtained from the Gene Expression Omnibus database under the accession number GSE227835. The dataset includes PBMC samples from patients with AChR antibody-positive myasthenia gravis.

Declaration of conflicting interests

The authors declare that there is no conflict of interest.

Ethics statement

This study was approved by the Ethics Committee of First Affiliated Hospital of Nanchang University. The approval number given by the ethical board is AF-SG-03-2.1-IIT.

Funding

This work was supported by the National Natural Science Foundation of China (82160252 and 82271439), the Natural Science Foundation of Jiangxi province (20224ACB206015), the double thousand talents program of Jiangxi province (jxsq2019101021), the Key Laboratory Project of Neurological Diseases in Jiangxi Province (2024SSY06072).

ORCID iD

Si Luo

Supplemental material

Supplemental material for this article is available online.

References

Gilhus

NE.

Myasthenia gravis. N Engl J Med 2016; 375: 2570–2581.

Gilhus

Skeie

Romi

, et al. Myasthenia gravis—autoantibody characteristics and their implications for therapy. Nat Rev Neurol 2016; 12: 259–268.

Gilhus

Verschuuren

JJ.

Myasthenia gravis: subgroup classification and therapeutic strategies. Lancet Neurol 2015; 14: 1023–1036.

Dalakas

MC.

Immunotherapy in myasthenia gravis in the era of biologics. Nat Rev Neurol 2019; 15: 113–124.

Thomas

Storey

RF.

The role of platelets in inflammation. Thromb Haemost 2015; 114: 449–458.

Gasparyan

Ayvazyan

Mukanova

, et al. The platelet-to-lymphocyte ratio as an inflammatory marker in rheumatic diseases. Ann Lab Med 2019; 39: 345–357.

Zhang

Wang

, et al. Effective immune-inﬂammation index for ulcerative colitis and activity assessments. World J Clin Cases 2021; 9: 334–343.

Weng

Zhu

, et al. Dynamic changes of peripheral inflammatory markers link with disease severity and predict short-term poor outcome of myasthenia gravis. J Neurochem 2024; 168: 2751–2761.

Huang

Wang

, et al. The systemic inflammation markers as possible indices for predicting respiratory failure and outcome in patients with myasthenia gravis. Ann Clin Transl Neurol 2023; 10: 98–110.

10.

Nehar-Belaid

Hong

Marches

, et al. Mapping systemic lupus erythematosus heterogeneity at the single-cell level. Nat Immunol 2020; 21: 1094–1106.

11.

Liu

Jin

, et al. Single-cell sequencing of immune cells from anticitrullinated peptide antibody positive and negative rheumatoid arthritis. Nat Commun 2021; 12: 4977.

12.

Wang

Zhang

Luo

, et al. Increased serum interleukin-41 correlates with disease severity in myasthenia gravis. Int Immunopharmacol 2024; 134: 112275.

13.

Koupenova

Livada

Morrell

CN.

Platelet and megakaryocyte roles in innate and adaptive immunity. Circ Res 2022; 130: 288–308.

14.

Vieira-de-Abreu

Campbell

Weyrich

, et al. Platelets: versatile effector cells in hemostasis, inflammation, and the immune continuum. Semin Immunopathol 2012; 34: 5–30.

15.

Boilard

Blanco

Nigrovic

PA.

Platelets: active players in the pathogenesis of arthritis and SLE. Nat Rev Rheumatol 2012; 8: 534–542.

16.

Liu

Gorzelanny

Schneider

SW.

Platelets in skin autoimmune diseases. Front Immunol 2019; 10: 1453.

17.

Rawish

Nording

Münte

, et al. Platelets as mediators of neuroinflammation and thrombosis. Front Immunol 2020; 11: 548631.

18.

Giorgiutti

Rottura

Korganow

, et al. CXCR4: from B-cell development to B cell-mediated diseases. Life Sci Alliance 2024; 7: e202302465.

19.

Kang

Bucala

The immunobiology of MIF: function, genetics and prospects for precision medicine. Nat Rev Rheumatol 2019; 15: 427–437.

20.

Ochando

Mulder

WJM

Madsen

, et al. Trained immunity—basic concepts and contributions to immunopathology. Nat Rev Nephrol 2023; 19: 23–37.

21.

Gurrea-Rubio

Fox

Castresana

JS.

CD6 in human disease. Cells 2025; 14: 272.

22.

Bayer

Nowak

O'Connor

KC.

Contribution of cellular immune dysregulation to myasthenia gravis pathology. Int Rev Neurobiol 2025; 182: 43–66.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB