Progress and challenges in the use of blood biomarkers in relapsing polychondritis

Summary

Relapsing polychondritis (RP) is a rare inflammatory disease with significant individual heterogeneity that involves systemic organs. The diagnosis of RP mainly depends on the clinical manifestations; currently, there are no molecular biomarkers routinely evaluated in clinical practice. Biomarkers have diagnostic or monitoring values and can predict response to treatment or the disease course. Over the years, many biomarkers have been proposed to facilitate diagnosis and prognosis. Unfortunately, ideal biomarkers to diagnose RP have not yet been discovered. Most of the molecular biomarkers in RP are immunological biomarkers, with autoantibodies and proteins related to cartilage damage in the blood being the most common. Alterations in some genes (HLA typing and UBA1 somatic mutation) were detected in patients with RP, which could serve as a potential biomarker for the diagnosis of RP. Moreover, proinflammatory cytokines and lymphocyte levels, and certain laboratory tests, have certain values of RP diagnosis and disease activity assessment but lack specificity and sensitivity. This review describes the different types of biomarkers and their clinical correlation with respect to the diagnosis of RP and disease activity. Research on biomarkers and disease pathology is ongoing to identify the ideal biomarkers that are sensitive and specific for RP.

This review summarizes different types of circulating biomarkers (including genes, autoantibodies, proteins, and immune cells, and so on) in patient with relapsing polychondritis (RP) and their clinical correlation with the disease diagnosis and disease activity, which would help understand disease pathophysiology and identify the ideal biomarkers that are sensitive and specific for RP.

Graphical Abstract

Graphical Abstract.

Introduction

Relapsing polychondritis (RP) is a rare systemic inflammatory disease characterized by the recurrent inflammation of organs and tissues containing cartilage structures and proteoglycan components. The main clinical manifestations are inflammation of the ear, nose, throat, trachea, and bronchi, and may also involve the cardiovascular system, joints, eyes, skin, and kidneys [1, 2]. RP has an insidious onset, strong heterogeneity, recurrent attacks, and remissions. Early diagnosis is difficult and the rate of misdiagnosis is high [3]. Clinically, about 30% of patients with RP show an overlap with other autoimmune diseases [4] such as vasculitis [5], rheumatoid arthritis (RA) [6], and systemic lupus erythematosus (SLE) [7], or hematological diseases such as myelodysplastic syndrome (MDS) [8]. In this review, we have discussed the epidemiology, etiology, pathogenesis, diagnostic criteria, assessment of disease activity, and clinical subgroups of RP. Besides, we have summarized all blood biomarkers and their clinical association in patients with RP, which can help understand the disease pathophysiology, effects of pharmacological intervention, and diagnostic value.

Epidemiology

As an orphan disease, RP has a low incidence of 0.71–4.5 per million people. The incidence is 3.5 per million people in the USA, 4 per million people in Japan, 4.5 per million people in the UK, 0.71 per million people in France, 2 per million people in Hungary, and 1–3.5 per million people across Central Europe. The age at diagnosis has been between 40 and 60 years, with a male-to-female ratio of almost 1:1. There were various degrees of delay in the diagnosis ranging from 12.6 to 22.8 months [9–12].

Etiology and pathogenesis

The etiology and pathogenesis of RP are currently unclear. The possible causative factors include infectious agents that share structural homology with self-antigens, such as heat shock protein (HSP) 60 of Mycobacterium tuberculosis and intracellular cytoskeletal and nuclear antigens. Reactive antibodies and B cells are induced by HSP60 and cross-react with disease-associated local self-antigens (cartilage), contributing to the pathogenic process by molecular mimicking, which activates the immune system [13]. In addition, chemical damage resulting from intravenous abuse of unknown drugs, the use of glucosamine chondroitin sulfate or hydralazine [14], or mechanical trauma such as perforation of the cartilaginous auricle can trigger the progression of RP by exposing recessive cartilage antigens [15–17].

In some genetically susceptible individuals (such as those with HLA-DR4/HLA-B/HLA-DRB1/HLA-DQB1/UBA1 somatic mutations) [18–21], the autoimmune system attacks self-antigens (mainly collagen, cartilage matrix protein); moreover, many inflammatory cells including lymphocytes, neutrophils, and macrophages are involved in cartilage damage. T lymphocytes, mainly CD4 T cells, are the inflammatory cells that infiltrate the cartilage matrix [22]. Under different conditions of stimulation, CD4 T cells differentiate into helper T cells 1 (Th1), Th2, Th17, and regulatory T cells (Treg) cells [23]. In patients with RP, there is an increase in interferon (IFN)-γ (secreted by Th1 cells and as macrophage-activating factor) [24, 25], macrophage inflammatory protein (MIP)-1, monocyte chemoattractant protein (MCP)-1 (recruiting and activating monocyte/macrophages), interleukin (IL)-8 (produced by activated monocyte/macrophages) [26], and a decrease in Th2 and Treg cell numbers [27]. A decrease in IL-10 released from Treg cells weakened the inhibitory effect on the secretion of inflammatory factors by monocytes/macrophages [28]. These findings show that RP is characterized by a Th1/Th2 imbalance and may thus be a Th1-associated disease with Th1-macrophage activation. In addition to the production of proinflammatory cytokines, there is an increase in the production of autoantibodies due to infiltration of the cartilage matrix by B lymphocytes and CD8 T cells, as well as CD8 T-cell-induced apoptosis [22, 29–31]. As the disease progresses, inflammatory cells invade the chondrocytes and there is an increase in the secretion of cathepsin K and cathepsin L; matrix metalloproteinase (MMP)-3 in chondrocytes; and MMP-1, 8, 9, and elastase in the perichondrium [22, 32, 33]. As the cartilage matrix degrades, the cartilage is gradually destroyed. Subsequently, the elastic tissue disappears resulting in cartilage deformities (Fig. 1).⁺⁺⁺⁺

Figure 1:

Pathogenesis of RP. The recruitment of inflammatory cells and cytokines/chemokines, production of autoantibodies, and secretion of proteolytic enzymes collectively lead to chondrocyte apoptosis and cartilage damage. Both cellular and humoral immunity are involved in this process. Abbreviation: COMP: cartilage oligomeric matrix protein; IL-18: interleukin-8; MCP-1: monocyte chemoattractant protein-1; MIP-1: macrophage inflammatory protein 1; MMP: matrix metalloproteinase; NR2B: glutamate receptor (GluR) ε2; Th1 cells: help T cells 1; Treg cells: regulatory T cells

Diagnostic criteria

The diagnosis of RP should be made based on a combination of clinical manifestations, laboratory examination, imaging studies, and histopathology findings. Currently, there is no unified diagnosis standard for RP; moreover, the early disease manifestations in some patients are not specific, which often leads to delayed diagnosis. McAdam [4], Damiani–Levine [34], and Michet [35] developed corresponding diagnostic criteria based on the clinical characteristics of 23, 10, and 112 patients, respectively, at a single center. Rose et al. applied three different diagnostic criteria in a retrospective analysis of 18 patients with RP and reported that the sensitivity of the Damiani and Levine criteria was 88.9% and that of the Michet and McAdam criteria were 66.7% and 50%, respectively. If the Michet criteria were modified to include ocular inflammation as the major criterion, and dermatologic and cardiovascular manifestations as the minor criteria, the sensitivity increased to 88.9% [36]. Furthermore, in another study in China that retrospectively analyzed 193 patients with RP and 130 control patients combined with other diseases, the sensitivity of the McAdam, Damiani–Levine, and Michet classification criteria were 72.0%, 89.1%, and 79.3%, respectively, and the specificity was 96.2%, 91.5%, and 96.2%, respectively. Polyarthritis, which was included as a minor criterion of the Michet criteria was a common clinical manifestation in Chinese patients with RP. When polyarthritis was included as the major criterion, the sensitivity of the modified Michet classification increased to 90.1% with little decrease in specificity (95.4%) [37].

Disease activity assessment

It is critical to assess disease activity and predict disease progression in patients with RP. However, there is a lack of adequate laboratory examinations to define active cartilage catabolism in patients with RP; moreover, there are no specific markers to monitor disease activity. In 2012, 27 international experts developed a standardized method to assess RP disease activity, the relapsing polychondritis disease activity index (RPDAI). RPDAI contains 27 items and each clinical manifestation is assigned a different value (ranging from 1 to 24 points). The corresponding scoring items and definitions are shown in Table 1. The scoring considers only those manifestations attributable to RP and that are present during past 28 days [38]. RPDAI scores were calculated online at https://qxmd.com/calculate/calculator_332/rpdai-relapsing-polychondritis-disease-activity-index. It is better if the RPDAI scores are independently determined by two experienced rheumatologists. Any disagreement between both can be resolved by a panel of rheumatologists who make the final decision [39]. The highest theoretical score is 265 points, which can evaluate disease activity and severity and predict the prognosis [38].

Table 1:

RPDAI scoring items [38]

Clinical manifestations	Definition	Score
Constitutional symptoms
Fever	>38°C or 100.4°F	2
Rheumatologic manifestations
Arthritis	Joint inflammation with soft tissue swelling or fluid in at least one joint	1
Chondritides
Manubriosternal chondritis	Pain or inflammation of the manubriosternal joint, lasting for >48 h (or of any duration if directly observed by a physician)	3
Sternoclavicular chondritis	Pain or inflammation of ≥1 sternoclavicular joint, lasting for a duration as listed	4
Costochondritis	Pain or inflammation of ≥1 costo-sternal joint, lasting for a duration as listed	4
Auricular chondritis	Unilateral or bilateral. Inflammation of the cartilaginous parts in the ear (helix, antihelix, tragus, or external auditory canal) sparing the lobule, lasting for a duration as listed	9
Nasal chondritis	Nasal pain or inflammation of the nasal cartilage, lasting for a duration as listed	9
Eye manifestations confirmed by an ophthalmologist
Episcleritis	Inflammation of the episclera	5
Scleritis	Inflammation of the sclera	9
Uveitis	Inflammation of the uveal tract (iris, ciliary body, choroid)	9
Corneal ulcer	Vascular leakage or sheathing of vessel walls	11
Retinal vasculitis	Corneal epithelial defect with underlying inflammation	14
Inflammatory data
C-reactive protein	>20 mg/L	3
Ear, nose, throat manifestations
Sensorineural deafness	Acute-onset deafness caused by lesions to the auditory nerve or cochlea, confirmed using audiometry	8
Vestibular dysfunction	Acute-onset vestibular involvement with vertigo or dizziness, nausea, and vomiting caused by lesions to the inner ear or vestibular nerve	12
Skin manifestations
Purpura	Cutaneous purpura	3
Renal manifestations
Hematuria	Hematuria ≥1+ on dipstick urinalysis; ≥1 × 104/ml	4
Proteinuria	Albuminuria > 1+ on dipstick or proteinuria >500 mg in a 24-h collection or albumin/creatinine ratio ≥30 mg/mmol or protein/creatinine ratio ≥50 mg/mmol	6
Renal failure	Impairment of renal function with increase in creatinine by >30% or a decrease in MDRD creatinine clearance >25%	17
Cardiovascular manifestations
Pericarditis	Pericardial pain or friction rub on clinical assessment or new onset of echographically confirmed pericardial effusion	9
Large and/or medium vascular involvement	Inflammation of large- or medium-sized vessels with supporting imaging abnormalities (such as parietal thickening, stenosis, and ectasia)	16
Myocarditis	Inflammation of the myocardium with raised cardiac enzymes or ECG changes without coronary obstruction	17
Acute aortic or mitral insufficiency	New aortic or mitral insufficiency	18
Neurologic manifestations
Motor or sensorimotor neuropathy	Peripheral neuropathy with clinically apparent motor deficit of at least one named nerve, with or without any sensory deficit, confirmed using electromyography with nerve conduction studies	12
Encephalitis	Diffuse or focal neuropsychological dysfunction due to inflammation of the brain parenchyma. Cerebral vasculitis should be recorded as large or medium vessel involvement	22
Respiratory manifestations: Laryngeal or tracheal or bronchial chondritis
Without acute respiratory failure	Respiratory chondritis: hoarseness, aphonia, cough, dyspnea, wheezing, inflammation with tenderness or swelling over laryngeal or tracheal cartilages	14
With acute respiratory failure	Acute respiratory failure: dyspnea due to acute airway obstruction from glottic, laryngeal, or subglottic inflammation requiring oxygen use or artificial ventilation	24

Causes other than RP should be excluded (e.g. trauma, infection) and stable persistent characteristics associated with damage are not evaluated [38].

Clinical subgroups

Currently, there are no clear international standards for the clinical classification of RP; however, three studies have described the clinical characteristics and subgroup classification of patients with RP in France, Japan, and the USA [8, 40, 41]. These three methods of grouping RP subgroups and corresponding clinical features are described in Table 2.

Table 2:

Three patterns of clinical classification in RP

Phenotypes	% (number)	Clinical features	References
Dion pattern
Hematologic phenotype	9% (12/142)	Age >55 years. MDS, cutaneous involvement, cardiac involvement, general symptoms, refractory disease common, tracheobronchial involvement rare. High rates of death (58%), serious infection (58%), and ICU admission (50%), and low remission rate (0)	[8]
Respiratory phenotype	26% (37/142)	Age ≤55 years. Tracheobronchial involvement common, may be present in combination with abnormal lung function. High rates of serious infection (35%), low death (13%), ICU admission (27%), and remission (0)
Mild phenotype	65% (93/142)	Age ≤55 years. Tracheobronchial involvement and MDS are rare. High remission and very low rates of death (4%), serious infection (6.5%), and ICU admission (2%)
Ferrada pattern
Type 1	14% (10/73)	Extensive cartilage damage in the ears, nose, upper respiratory tract: ear chondritis (100%), TM (100%), saddle nose deformity (90%), SGS (80%). Shortest diagnosis time, youngest age of diagnosis, highest activity	[40]
Type 2	29% (21/73)	Lower respiratory symptoms are predominant: TM (100%), BM (52%), tenosynovitis/synovitis (52%), ear chondritis (43%), hearing loss (33%), inflammatory eye disease (19%); in the absence of saddle nose deformity and SGS. Longest delay to diagnosis and the highest percentage of work disability. Negatively correlated with ear chondritis
Type 3	57% (42/73)	Minimal overt cartilage damage: tenosynovitis/synovitis (60%), ear chondritis (55%), inflammatory eye disease (26%), hearing loss (19%), skin involvement (14%), SGS (7%), saddle nose deformity (2%), no TM or BM. High disability (26%), hearing loss (19%), and active disease
Shimizu pattern
A subgroup	49.4% (118/239)	Auricular involvement without respiratory involvement, conjunctivitis, non-erosive arthritis, CNS involvement	[41]
R subgroup	19.7% (47/239)	Respiratory involvement without auricular involvement, saddle nose deformity, progressive disease course
O subgroup	29.3% (70/239)	Auricular and respiratory involvement, high MMP3 levels, cardiovascular involvement, longer disease duration, saddle nose deformity, progressive disease course

Based on the clinical manifestations, disease progression, and prognosis, Dion et al. divided RP into three different clinical phenotypes, namely, hematologic, respiratory, and mild. The factors associated with death were male sex, cardiac abnormalities, and concomitant MDS or other hematologic malignancies. Hematologic phenotype usually had concomitant MDS, contributing to poor prognosis. These patients were mainly older men who exhibited strong cutaneous involvement (neutrophilic dermatoses), cardiac involvement, general symptoms, and refractory disease. Tracheobronchial involvement is rare. Tracheobronchial involvement is a hallmark of respiratory phenotype in which cutaneous involvement and auricular chondritis are less frequent. These patients are prone to infection and have an intermediate prognosis. The mild phenotype, which has the largest population and a low number of RP manifestations and treatments, is associated with the absence of severe complications and possible disease remission [8].

In addition, Ferrada et al. defined three subgroups based on eight clinical variables (saddle nose deformity, subglottic stenosis [SGS], tracheomalacia [TM], bronchomalacia [BM], ear chondritis, tenosynovitis/synovitis, inflammatory eye disease, and audiovestibular disease) in a prospective observational cohort, including patients with Type 1 RP, Type 2 RP, and Type 3 RP. Patients with Type 1 RP exhibited a classic disease pattern with ear chondritis and extensive damage to the cartilage of the nose and upper airway. These patients had the shortest time to diagnosis, the youngest age at diagnosis, the highest activity levels, and more pronounced weight loss and wheezing. Type 2 RP is similar to the respiratory phenotype in the study by Dion. The clinical manifestations of Type 2 RP were mainly lower respiratory symptoms with genital ulcers and wheezing. Patients with predominant lower airway involvement had the lowest incidence of ear involvement. A prior diagnosis of asthma was significantly more common in patients in the Type 2 RP subgroup, who experienced an average delay of 10 years in diagnosis. Patients with Type 3 RP, similar to the mild phenotype in Dion's subgroups, lacked features of overt cartilage damage. They also exhibited a high frequency of disability, hearing loss, and active disease, and skin involvement was more obvious than that in the other subgroups [40].

Shimizu et al. observed a strong inverse relationship between respiratory and auricular involvement. Based on the patterns of clinical manifestations, researchers classified patients into one of the following subgroups: respiratory involvement but without auricular involvement (R subgroup), auricular involvement but without respiratory involvement (A subgroup), and auricular and respiratory involvement (O subgroup). Respiratory complications, which were frequent in subgroup R were the most common cause of death (55% of total deaths). In addition to the above clinical manifestations, saddle nose deformity and progressive disease course were also often observed in the R subgroup. Nasal involvement was significantly related to respiratory involvement, and cardiovascular involvement was not observed in the R subgroup. Conjunctivitis, non-erosive arthritis, and central nervous system (CNS) involvement were more prevalent in A subgroup. Eye involvement was positively associated with joint involvement, as was auricular involvement with CNS involvement. CNS involvement was another major prognostic factor for RP. Apart from saddle nose deformity and progressive disease course, subgroup O was characterized by cardiovascular involvement, high MMP3 levels, and longer disease duration, in which cardiovascular involvement was positively associated with auricular, CNS, and renal involvement. Further, subgroups with respiratory involvement were more likely to receive treatment with biological agents such as infliximab [41].

As a systemic inflammatory disease, RP is known for widespread organ involvement and various clinical manifestations. Subgroup analysis facilitates better recognition, timely diagnosis and treatment, decreasing uncontrolled inflammation and ultimate organ damage. For example, different phenotypes or subgroups may be associated with different pathophysiologic mechanisms, involving a paraneoplastic process in the hematologic phenotype and immunization against tracheobronchial cartilage antigens in the respiratory phenotype. The three studies discussed above found a mutually exclusive relationship between respiratory involvement and ear involvement, illustrating the similarities between RP in different countries and regions. In addition, researchers showed that patients with respiratory involvement were more susceptible to infection than those without, which can lead to airway obstruction and subsequently respiratory failure, the main cause of death [10], patients in these subgroups also received more aggressive treatments to improve their prognosis outcomes. Subgroup analysis considers the heterogeneity of RP, which is beneficial in the therapy and management of patients. However, the current subgroups are clinically relevant and lack of molecular typing subgroups (such as anticitrullinated peptide antibody-positive and -negative RA) [42] to explore disease pathogenesis and achieve precision therapy. Future studies should investigate whether subgroups of patients with RP also differ in terms of causal factors such as genetic risk or environmental exposure, immunologic mechanisms of diseases, and responses to treatment.

Blood biomarkers

Diagnosis of RP and disease activity monitoring often relies on clinical signs and symptoms, nonspecific inflammatory marker levels such as those of C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), and auxiliary imaging modalities such as laryngoscopy, 99mTc-MDP imaging, and computed tomography. Diagnosing and objectively monitoring RP disease activity is challenging. Thus, early detection and treatment are beneficial in the prognosis of RP. Blood analysis is a noninvasive diagnostic method widely used in clinical practice. In the subsequent sections, we have summarized studies related to the evaluation of blood biomarkers in patients with RP (Table 3) from the perspectives of genes, autoantibodies, proteins, cytokines, immune cells, and laboratory examination to provide possible directions for clinical diagnosis and follow-up studies.

Table 3:

Blood biomarkers in patients with RP

Biomarkers	No. Of patients	Clinical relevance	Methods	Country	References
Genes
HLA-DR4	41	Prevalent in patients with RP (56.1%) than the HCs (15.5%)	PCR	Germany	[18]
HLA-DR6	62	Negatively correlated with the degree of organ involvement	NA	Germany	[7]
HLA-DRB1*16:02	204	Prevalent in patients with RP (3.9%) than in the HCs (0.6%)	WAKFlow system	Japan	[19]
HLA-DQB1*05:02	204	Prevalent in patients with RP (7.4%) than in the HCs (2.2%)	WAKFlow system	Japan	[19]
HLA-B*67:01	204	Prevalent in patients with RP (3.9%) than in the HCs (0.95%)	WAKFlow system	Japan	[19]
UBA1 variant	14	High prevalence in 73% of men and correlation with skin involvement	Sanger sequencing, ddPCR, PNA-clamping PCR	Japan	[20]
	92	13.27% prevalence in all patients with RP and 50% prevalence in men; correlation with old age, fever, ear chondritis, skin involvement, DVT, pulmonary infiltrates, high mortality, acute-phase reactants, and hematologic abnormalities.	Whole-exome sequencing	USA	[21]
IL1β mRNA	30	Upregulated in PBMCs stimulated with mitogens to reflect disease activity	RT-PCR	Japan	[48, 49]
IL-10 mRNA	22	Upregulated in PBMCs of inactive patients with RP because of propionate-producing bacteria	RT-PCR	Japan	[49]
Foxp3 mRNA	22	Compensatively upregulated in PBMCs stimulated with mitogens due to Treg cells inhibition to reflect disease activity
Autoantibodies
Anti-cartilage antibody	9	High titers correlate with active RP; decreases with therapy or clinical remission	IIF	UK	[30]
Anti-native CⅡ antibody	15	Positive in acute RP, correlated with RP severity and treatment	IIF	USA	[29]
	6	Upregulated in RP and helps distinguish patients with RP from the HCs	ELISA	Germany	[61]
	97	Higher IgG titers in patients with RP than in the HCs	ELISA	Sweden	[31]
Anti-native CⅨ antibody	6	Upregulated in RP and helps distinguish patients with RP from the HCs	ELISA	Germany	[61]
	97	21% positive in RP	ELISA	Sweden	[31]
Anti-native CXI antibody	97	21% positive in RP	ELISA	Sweden	[31]
Anti-matrilin-1 antibody	97	Upregulated IgG and IgM titers in 13% of patients with RP and correlated with respiratory symptoms	ELISA	Sweden	[31]
Anti-COMP antibody	97	Upregulated IgG and IgM titers in 14% of patients with RP and helps distinguish patients with RP from those with WG and the HCs	ELISA	Sweden	[31]
Anti-NR2B antibody	1	Positive in the CSF and serum of a patient with limbic encephalitis associated with RP	NA	Japan	[64]
ANA	111	Positive in 27% of patients with RP; 10% higher titers found in RP combined with other disorders	IIF	USA	[65]
ANCA	33	Positive in 24% of patients with RP and correlated with active disease	IIF	France	[66]
	1	Positive and upregulated due to the drug PTU	Immunological tests	China	[67]
Cytokines and cytokine receptors
sTREM-1	15	Upregulated in patients with RP and active RP, correlated with treatment and clinical course	ELISA	Japan	[24]
IFN-γ	15	Upregulated in patients with RP with AUC 0.77	ELISA	Japan	[24]
CCL4	15	Upregulated in patients with RP with AUC 0.79	ELISA	Japan	[24]
VEGF	15	Upregulated in patients with RP with AUC 0.78	ELISA	Japan	[24]
MIF	5	Higher than the HCs but no difference between patients with RP and WG	ELISA	Japan	[69]
MCP-1, MIP-1, IL-8	22	Upregulated in patients with active RP, suggestive of monocyte/macrophage activation	ELISA	USA	[26]
Protein biomarkers
COMP	21	Elevated in the active stage and unaffected by treatment	ELISA	France	[70]
	15	Elevated in RP but cannot distinguish between patients with active and inactive RP	ELISA	Japan	[24]
	1	Decreased in RP and increased after treatment	RIA	Sweden	[71]
CMP	1	Elevated in patient with symptoms pronounce	RIA	Sweden	[71]
MMP3	15	Upregulated in patients with RP with AUC 0.80	ELISA	Japan	[24]
	22	Upregulated in RP and correlated with respiratory involvement	ELISA	Japan	[41]
	30	Upregulated in RP; correlated with respiratory involvement; positively correlated with RPDAI in patients without respiratory involvement	ELISA	Japan	[48]
uTIINE	1	Upregulated in RP and associated with treatment	ECL	USA	[75]
Lymphocyte subsets
NKT cells	10	Decreased in patients with RP and led to abnormal immune function	Flow cytometry	Japan	[76]
CD4 T cells+	10	Increased in patients with RP, suggestive of Th1 cell activation		Japan	[76]
CD4 T cells+	19	Treg and Th2 cells decreased in patients with RP, leading to weakening of immunosuppressive function and Th1/Th2 imbalance		China	[27]
Other routine markers
Laboratory examinations	170	White blood cell, neutrophil, monocyte, and platelet counts increased in RP; CAR, NLR, PLR, and neutrophils positively correlated with RPDAI	NA	China	[39]
CICs	15	Positive in 20% of patients with RP	NA	USA	[29]

Genetic biomarkers

Human leukocyte antigen

Human leukocyte antigen (HLA) is closely related to the immune system. Focusing on HLA loci can not only help explore disease pathogenesis but also serve as important biomarkers for clinical diagnosis. Animal experiments have been used to show associations between RP and HLA. HLA-DQ6/8 double transgenic mice develop auricular chondritis resembling RP [43], supporting the involvement of antigen recognition through HLA molecules in the pathophysiology of RP. The relationship between HLA alleles and RP was found for the first time in German Caucasians. The frequency of DR4 antigen in patients with RP (23/41, 56.1%) was significantly higher than that in healthy controls (HCs) (52/204, 15.5%), indicating that HLA-DR4 was significantly associated with RP in the German population. However, there was no predominance of any DR4 subtype allele in patients with RP and no differences in MHC class I antigen frequencies between patients with RP and the HCs [18]. Another study also confirmed the above conclusions and reported that the degree of organ involvement was negatively correlated with HLA-DR6 [7]. Günaydin et al. reported that 85.71% (6/7) of patients with RP with articular involvement were positive for HLA-DR4 [44]. Rajaee et al. reported a case of RP combined with RA showing the presence of HLA-DR4 [45], suggesting that HLA-DR4 was probably related to arthropathy in patients with RP; however, large cohorts are needed to verify whether HLA-DR4 is a risk factor for joint involvement. However, no correlation was observed between HLA-DR4 and RP in the Japanese population. HLA-DRB1*16:02, HLA-DQB1*05:02, and HLA-B*67:01 were found to be significantly associated with RP (OR = 6.76 [95% CI: 2.73, 16.74], OR = 3.53 [95% CI: 1.93, 6.46] and OR = 4.26 [95% CI: 1.84, 9.85], respectively) and showed the existence of weak linkage disequilibrium with each other. The genetic characteristics of patients with RP were distinct from those with rheumatic diseases, such as RA, SLE, Behçet disease, and Takayasu arterifis. However, the positive incidence of the above three HLA loci was low (8/204, 15/204, 8/204, respectively), limiting their usage as a marker for diagnosing RP [19]. Further, it is useful to expand the number of RP subjects for identifying the clinical characteristics of patients carrying these HLA alleles.

Somatic UBA1 variants

Somatic variants in the ubiquitin-activating enzyme-1 (UBA1) gene were discovered in individuals with systemic inflammation of the cartilage, skin, and blood vessels, accompanied by hematological abnormalities namely the vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic (VEXAS) syndrome, which only affects male patients. It has been reported that 60% of patients with VEXAS had an RP phenotype [46]. UBA1 was detected in 13 of the 14 Japanese patients with RP; 73% (8/11) of male patients had somatic UBA1 variants (c.121A>C, c.121A>G, or c.122T>C, resulting in p.Met41Leu, p.Met41Val, or p.Met41Thr, respectively). Patients who were UBA1 variant–positive showed a significantly high prevalence of skin involvement [20]. Ferrada et al. have reported the positive rate of somatic UBA1 variants as 13.27% (13/98) and 50% (13/26) in American patients with RP and male patients, respectively. Patients with UBA1 variant-positive were all male, elderly (≥45 years), and usually presented with fever, ear chondritis, skin involvement, deep vein thrombosis (DVT), pulmonary infiltrates, and rarely airway chondritis or costochondritis at disease onset. In addition, the UBA1 variant was associated with high mortality and acute-phase reactants as well as hematologic abnormalities (high mean corpuscular volume [MCV]; low platelet, hemoglobin, lymphocyte, and monocyte counts; MDS; and multiple myeloma). Male sex, MCV >100 fL, and platelet count <200 × 10³/μL can be used to differentiate VEXAS-RP (patients with UBA1 variant-positive) from RP with 100% sensitivity and 96% specificity [21].

Single nucleotide polymorphisms (SNPs)

There were nine same SNPs (RNF207, COL22A1, GPAA1, RECQL4, FLCN, LIG3, CCDC61, PCP2, and MYH15 gene mutations) in two patients with RP (the proband and the proband's mother) in a Chinese family, demonstrating that coinheritance of multigene mutation may predispose to RP. Besides, six of the nine SNPs (RNF207, COL22A1, GPAA1, RECQL4, FLCN, and LIG3) were mutated in the 2-year-old daughter of the proband. However, this study did not detect the prevalence of these nine SNPs in a large cohort population. Therefore, in-depth research is required to explore SNPs in patients with RP and to provide more useful genetic markers for clinical diagnosis and treatment [47].

RNA biomarkers

The expression of cytokines IL1β, IL6, and TNFα mRNA was found to be significantly lower in freshly isolated peripheral blood mononuclear cells (PBMCs) of patients with RP at clinically stable compared with the HCs. When PBMCs were stimulated with mitogens to mimic disease activation, IL1β mRNA increased significantly in patients with RP compared with the HCs. IL6 was also upregulated but the increase was not statistically significant. TNFα was maintained at a low level in patients with RP. Furthermore, the level of IL1β mRNA in patients with respiratory involvements after stimulation was positively correlated with MMP-3 (a marker indicative of respiratory manifestations in patients with RP). These data suggest that IL1β can be used as an indicator to identify patients with RP at the active stage, and immune responses mediated by IL1β have certain effects on bronchial inflammation in RP [48]. However, these findings need to be verified in a real-world setting rather than merely simulating disease relapse or activity at the cellular level. In another study, these authors reported that propionate-producing gut microbes were predominant in patients with RP [49]. In the intestine, propionate induced the accumulation of IL10-producing Treg cells [50], which led to higher IL10 mRNA levels in the PBMCs of patients with RP than in the PBMCs obtained from the HCs [49]. However, IL1β, IL6, and TNFα mRNAs associated with proinflammation were lower in RP patients than in the HCs. After stimulation with mitogens, these cytokines showed an opposite trend wherein IL10 decreased and IL1β, IL6, and TNFα increased [49]. Downregulation of IL10 resulted from T-cell hyporesponsiveness against mitogen stimulation in patients with inflammatory diseases [51, 52]; however, the mRNA level of Foxp3, a transcription factor required for Treg cell function [53], might compensatively be increased in the PBMCs of patients with RP [49]. Upon stimulation with mitogens, IL6 and TNF-α levels were not significantly elevated in the PBMCs of patients with RP compared with those of the HCs; thus, high mRNA levels of IL10, IL1β, and Foxp3 were used to distinguish between inactive and active patients. A decrease in IL10 mRNA by Treg cells and an increase in inflammatory cytokine IL1β mRNA by PBMCs might lead to chondritis in patients with RP [48, 49].

Autoantibodies

Anti-cartilage antibody

Cartilage can be categorized into hyaline cartilage, elastic cartilage, and fibrocartilage, and is composed of cartilage tissue and its surrounding perichondrium. Cartilage tissue is composed of chondrocytes, matrix, and fibers [54]. Hyaloid cartilage contains various types of collagen (type Ⅱ collagen [CⅡ], CⅨ, CXI, CⅥ, CⅢ, CⅩ, CXII, and CIV) [55], proteoglycans [56] and non-collagen proteins (cartilage oligomeric matrix protein [COMP], cartilage matrix protein matrilin-1, etc.) [57, 58], which are found on the articular surface of the bones of synovial joints, ribs, nose, trachea, bronchi, larynx, and growth plates [54]. The outer ear, larynx, and epiglottis show the presence of elastic cartilage in which the extracellular matrix (ECM) is mainly composed of CⅡ, proteoglycan, and elastin fibers [59]. Fibrocartilage contains CⅠ and CⅥ, and, to a lesser extent, CⅡ and proteoglycans [59]. It is predominant in the intervertebral discs, meniscus, bone–tendon junctions, and ligament–tendon junctions [54]. Based on the above data, it is clear that the mainly hyaline cartilage and elastic cartilage are involved in RP. Proinflammatory cells infiltrate cartilage tissues, leading to the high turnover of some components of the cartilage matrix such as collagens, matrilin-1, COMP, and the proteoglycan aggrecan. Therefore, specific autoantibodies against collagens, matrilin-1, and COMP are produced.

Anti-cartilage antibodies were first identified in 1973 in the serum of a patient with RP by indirect immunofluorescence (IIF) using rat or bovine costal cartilage sections [60]. Subsequently, using IIF, six out of nine patients with RP demonstrated the presence of anti-cartilage antibodies, which were predominantly IgG that varied in titers from 1:1 to 1:320. Higher titers were found during the active phase of the disease and tended to decrease upon steroid therapy or after clinical remission [30].

Anti-collagen antibody

Foidart et al. found antibodies to native CⅡ in the serum of patients with acute-phase RP. Antibody titers were related to disease severity. The titers decreased significantly after treatment with prednisone and progressive clinical improvement [29]. In 1991, anti-native and denatured CIX autoantibodies and anti-denatured CXI autoantibodies were discovered in patients with RP. Anti-native CⅡ and CIX antibodies could distinguish patients with RP from the HCs [61]. In another study, IgG titers of anti-CII in patients with RP were highly significantly different compared with the HCs. Moreover, 21% of patients with RP were found to be anti-CIX and CXI positive, but there were no significant differences between patients with RP and the HCs [31].

Anti-matrilin-1 antibody

Matrilin-1 is abundant in the hyaline cartilage and can mediate interactions between the macromolecular components of the ECM, e.g. collagens and proteoglycans [62]. Elevated positive anti-matrilin-1 IgG and IgM titers were found in 13% (13/97) of patients with RP and the levels were positively correlated with respiratory symptoms. In addition, titers > 0 were identified in 24% (23/97) of patients with RP. Interestingly, some patients with high anti-matrilin-1 antibody titers also had very high anti-CIX, anti-CXI, and anti-COMP titers. On the other hand, relatively low titers of anti-matrilin-1 were detected in a few patients or were even absent in patients with autoimmune diseases such as Wegener's granulomatosis (WG), RA, and osteoarthritis (OA) [31].

Anti-COMP antibody

COMP is a large pentameric glycoprotein that interacts with several ECM proteins in cartilages and other tissues; it plays a role in collagen secretion and fibrillation, chondrocyte proliferation, and conferring mechanical strength to tendons [63]. Elevated antibody titers to COMP were found in 14% (14/97) of patients with RP. COMP-specific IgG and IgM titers in patients with RP were significantly higher than those in patients with WG. However, anti-COMP antibody titers did not differ between patients with RP and those with other disorders such as RA, SLE, and OA. In addition, anti-COMP antibody titers in patients with RP were higher than those in the HCs when only analyzing individuals having titers >0 [31].

Anti-NR2B antibody

Anti-glutamate receptor (GluR) ε2 (NR2B) autoantibodies IgG and IgM antibodies in the cerebrospinal fluid (CSF) and IgG antibodies in serum were found in a patient with limbic encephalitis associated with RP. These antibodies were not detected in other body fluids [64]. This patient was previously diagnosed with aseptic meningitis and did not undergo any specific immunosuppressant therapy, which led to recurrent and worsening neurological symptoms. Therefore, it is necessary to detect anti-GluRε2 antibodies in a larger cohort to determine their diagnostic value for the early diagnosis of RP. However, validation may be difficult considering the low incidence of RP and the involvement of the nervous system only in 3% of the cases of RP [32].

Non-specific antibodies

In addition to the aforementioned autoantibodies, low levels of antinuclear antibodies (ANA) were present in patients with RP. Piette et al. reported that 27% of patients with RP were ANA positive, of which 18% and 9% (10/111) had low and high ANA titers, respectively. Furthermore, among those 10 patients, 5 had Sjögren syndrome (SS) and 2 had MDS, suggesting that ANA was not a specific antibody for RP. A significantly high ANA titer in a patient with RP strongly suggested the presence of other overlapping disorders [65].

In a study by Papo et al., 24% (8/33) of patients with RP were positive for antineutrophil cytoplasmic antibodies (ANCA). The study reported that 36.84% (7/19) of active patients with RP were ANCA-positive and only 7.14% (1/14) of inactive patients were ANCA-positive indicating that ANCA (either diffuse or perinuclear) might tend to be present in the active disease stage [66]. In a patient with RP who also had Graves’ disease (GD), ANCA positivity was caused by the drug propylthiouracil (PTU) that was used to treat GD. When PTU was replaced by iodine-131 therapy, the symptoms resolved rapidly, the inflammatory marker (CRP and ESR) levels returned to normal, and myeloperoxidase (MPO)-ANCA levels decreased. Thus, PTU should be avoided in patients with RP who are also affected by GD, especially when they are positive for ANCA [67].

Cytokines and cytokine receptors

Triggering receptor expressed on myeloid cells (TREM)-1 is a type I transmembrane receptor of the immunoglobulin superfamily [68]. The soluble form of TREM-1 (sTREM-1) is upregulated in the serum of patients with RP and can be useful in differentiating patients with RP from the HCs with high sensitivity and specificity (area under the ROC curve [AUC]: 0.90). A sTREM-1 cutoff value of 158 pg/ml had a sensitivity of 86.7% and specificity of 86.7%. Furthermore, patients with active RP had higher sTREM-1 levels compared with those with inactive RP. In a patient with active RP who was treated with prednisolone and MTX, the time required for the increase in sTREM-1 levels was associated with the clinical course. Serum sTREM-1 levels decreased when clinical symptoms were relieved and increased when the disease recurred. However, sTREM-1 levels increased by varying degrees in RA, SLE, primary Sjögren syndrome, and HTLV-1-associated myelopathy, suggesting that sTREM-1 was suitable as a disease-activity marker but not as a diagnostic marker for RP [24].

Serum samples of patients with RP showed significantly higher IFN-γ, CCL4, and VEGF levels compared with samples from the HCs, and the AUCs were 0.77, 0.79, and 0.78, respectively [24]. As a proinflammatory cytokine, serum macrophage migration inhibitory factor (MIF) levels in patients with RP were significantly higher than those in the HCs, but there was no significant difference in MIF levels between patients with RP and WG [69]. Besides, MCP-1, MIP-1, and IL-8 levels were much higher in patients with active RP compared with the HCs. These results were consistent with previous reports on MIF levels. All three cytokines that were elevated in patients with RP were proinflammatory chemokines involved in regulating monocyte/macrophage function, which suggests that macrophages are activated in patients with RP [26].

Protein biomarkers

COMP is an ECM glycoprotein expressed in the cartilage, fibroblasts, tendons, ligaments, synovium, vascular smooth cells, and myofibroblasts [63]. Infiltration of cartilage tissue by inflammatory cells destroys the cartilage matrix, which, in turn, results in the release of COMP into the blood. Lekpa et al. have determined serum COMP levels in patients with RP in the active and inactive stages and found that COMP was elevated in the active stage. Moreover, treatment with steroids did not alter COMP levels; thus, it could be used for monitoring cartilage damage in RP [70]. However, in another study, although serum COMP tended to be higher in patients with RP than in the HCs, and in patients with active RP than in those with inactive RP, the differences were not statistically significant [24]. Serum COMP levels showed an opposite trend in a case report: patients with RP had lower COMP levels than the HCs before treatment, and the levels returned to normal after treatment [71]. This finding could be explained by different disease conditions of RP, sample sizes, and measurement methods. In this case study, CMP, which was mainly expressed in tracheal cartilage but was not detectable in mature articular cartilage, was present in high levels as the symptoms of patients with RP were pronounced. Thus, CMP release is indicative of catabolic processes of the cartilage [71].

The MMP family is involved in the breakdown of ECM proteins during tissue remodeling in conditions such as arthritis and during tumor metastasis. MMP-3 can degrade CII, CIII, CIV, CIX, and CX, as well as proteoglycans, fibronectin, laminin, and elastin. Moreover, it can activate other MMPs, which is important for connective tissue remodeling [72, 73]. Serum MMP3 levels were higher in patients with RP (AUC = 0.80) compared with the HCs [24]. A study by Shimizu reports an inverse relationship between the incidence of respiratory involvement and the incidence of auricular involvement in RP. MMP3 was significantly higher in patients with respiratory involvement compared with the HCs, but not in patients with auricular involvement. Moreover, MMP3 concentration in both respiratory and auricular involvement was higher than that in either respiratory or auricular involvement [41]. Another study confirmed these conclusions, and serum MMP3 levels in patients with respiratory involvement were higher than in those without respiratory involvement. Besides, there is a significant positive correlation between the RPDAI score and serum MMP3 levels in patients without respiratory involvement, suggesting that MMP3 might be used to monitor disease activity in these patients [48].

CII is highly expressed in cartilage tissue and can be cleaved by MMPs-1, -8, and -13 to generate a new type-Ⅱ collagen neoepitope (TⅡNE), which has high renal clearance and can be detected in the urine [74]. In a case report, urine TⅡNE (uTⅡNE) levels were associated with treatment with the TNF inhibitor etanercept. After medication, uTⅡNE levels returned to normal; however, uTⅡNE levels increased suddenly once the therapy was stopped. Inhibition of CII degradation was consistent with the use of TNF inhibitors, suggesting that uTⅡNE levels can reflect RP severity and aid clinical decisions in adjusting steroid and other immunosuppressant therapies. However, this indicator should be studied in a wider population of patients with RP [75].

Lymphocyte subsets

The study by Takagi found that natural killer T (NKT) cell levels are decreased in patients with RP [76]. NKT cells play several roles in autoimmune liver disease, including driving anti- and pro-inflammatory responses and regulating other types of immune regulatory cells functioning as Treg cells [77]. Quantitative and functional deficiencies in NKT cells have been reported in SLE [78]. A decrease in NKT cells leads to abnormal immune function and promotes the progression of RP. In addition, the IFN-γ/IL-4 ratio in CD4 T cells in patients with RP was found to be significantly higher than the HCs [76], which was consistent with the aforementioned high concentrations of IFN-γ in patients with RP [24]. The absolute value and percentage of Treg cells were decreased [27]. In accordance with the findings mentioned above, IL-10 (secreted by Treg cells) mRNA was downregulated in the PBMCs of patients with RP after mitogen stimulation [49]. Treg imbalance weakens immunosuppressive function and promoted the occurrence of RP. Moreover, the absolute numbers of Th2 cells were also reduced in patients with RP. Although the Th1/Th2 ratio did not increase significantly in patients with RP compared with the HCs (13.51 in patients with RP vs 10.33 in the HCs) [27], RP could be considered a Th1-biased autoimmune disease as illustrated in the Etiology and Pathogenesis section.⁺⁺⁺

Other routine markers

Cao et al. performed routine and simple hematological tests and found that albumin and hemoglobin levels were decreased, and fibrinogen levels and the counts of white blood cells, neutrophils, monocytes, and platelets were increased in patients with RP compared with the HCs. The CRP to albumin ratio (CAR), neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR), and neutrophil counts were positively correlated with RPDAI, whereas albumin, hemoglobin, and lymphocytes were negatively correlated. Three new inflammatory markers, namely, CAR, NLR, and PLR can be used as markers to determine the disease activity of RP [39]. In addition, vitamin D insufficiency was found to be common in patients with RP; 91% of the 22 patients with RP had vitamin D insufficiency and 18% had vitamin D deficiency, and vitamin D levels were negatively correlated with ESR [79]. Low levels of circulating immune complexes (CICs) were detected in the serum of 3 of 15 patients with RP. CIC formation might be secondary to the release of collagen fragments into the circulation following tissue injury, which can explain some of the clinical manifestations that do not involve cartilage structures [29].

Conclusions

Biomarkers play an important role in improving the diagnosis, treatment, and prognosis of patients with diseases. Compared with cartilage biopsy or nonspecific inflammatory markers such as CRP and ESR, the emergence of other blood markers not only broadens the horizons to further understand RP but also provides new ideas for the diagnosis and treatment of RP. In this review, we have summarized several circulating biomarkers that cover aspects such as genes, autoantibodies, proteins, cytokines, and immune cells, as well as laboratory examinations of blood, urine, and CSF and their clinical associations in patients with RP. These evaluations can help us understand disease pathophysiology and the effects of pharmacological intervention. From the perspective of the most specific and sensitive biomarker, a sTREM-1 cut-off value of 158 pg/ml had a good diagnostic performance with a sensitivity of 86.7% and specificity of 86.7% and area under the ROC curve of 0.90, and it could distinguish between active and inactive patients with RP [24]. From the perspective of large sample size and multiple multicenter validation, HLA-DRB1*16:02, HLA-DQB1*05:02, and HLA-B*67:01 were the risk factors of RP [19]. A high prevalence of UBA1 variants was present in male patients with RP [20, 21]. The role of anti-CII antibodies in the diagnosis of RP was validated by multiple cohorts from different countries [29, 31, 61]. Anti-matrilin-1 antibody [31] and MMP3 [24, 41, 48] were found to be associated with respiratory symptoms in a large number of patients with RP. These biomarkers show potential for use in a clinical setting. Although the circulating biomarkers reviewed in this study could be useful in the diagnosis and management of RP, significant work remains to be done. The use of new blood biomarkers and establishment of biomarkers reference intervals need validation and improvement based on clinical data. Moreover, due to the rarity of the disease, the heterogeneity of genetic characteristics, and clinical manifestations of patients with RP in different regions, no blood biomarker is currently used in a clinical setting as a diagnostic criterion for RP. Additionally, related research on blood biomarkers to determine susceptibility factors, pathogenesis, and disease progression of RP is lacking. Therefore, in-depth studies in a clinical setting are warranted prior to the use of blood biomarkers for the diagnosis of RP and to determine disease activity and prognosis.

Currently, high-throughput technologies in the field of proteins (proteomics), genes (genomics), RNA (transcriptomics), and metabolites (metabolomics) can reveal interaction networks at the molecular level and help identify novel biomarkers and treatment targets, thereby providing new opportunities to understand the pathophysiology of complex diseases [80]. Nine novel autoantibodies have been identified that can be used for the diagnosis of ACPA-negative RA by using high-throughput HuProt microarrays [81]. Genome-wide association studies have identified three novel coding variants and four novel susceptible gene regions for SLE in the Han population, providing new insights into the biological mechanisms of SLE [82]. Moreover, urine metabolomics studies have reported non-invasive biomarkers that reflect subtle metabolic discrepancies in response to specific diseases or therapeutic interventions [83]. As there is inadequate omics research for RP, future studies can focus on using genomics, proteomics, metabolomics, transcriptomics, lipidomics, glycomics, and other technologies to identify novel biomarkers related to RP. To summarize, with the expansion of RP-related research and a better understanding of RP, breakthroughs can be made in the prevention, clinical diagnosis, and treatment of RP.

Abbreviation

ANA

antinuclear antibodies

ANCA

antineutrophil cytoplasmic antibodies

BM

bronchomalacia

CICs

circulating immune complexes

COMP

cartilage oligomeric matrix protein

CNS

central nervous system

CSF

cerebrospinal fluid

C II

type II collagen

CAR

CRP to albumin ratio

DVT

deep vein thrombosis

ECM

extracellular matrix

GD

Graves disease

HCs

healthy controls

HLA

human leucocyte antigen

HSP60

heat shock protein 60

IIF

indirect immunofluorescence

IL-18

interleukin-8

MCP-1

monocyte chemoattractant protein-1

MCV

mean corpuscular volume

MDS

myelodysplastic syndrome

MIF

macrophage migration inhibitory factor

MIP-1

macrophage inflammatory protein 1

MMP

matrix metalloproteinase

MPO

myeloperoxidase

NKT cells

natural killer T cells

NR2B

glutamate receptor (GluR) ε2

NLR

neutrophil to lymphocyte ratio

OA

osteoarthritis

PSMCs

peripheral blood mononuclear cells

PTU

propylthiouracil

PLR

platelet to lymphocyte ratio

RA

rheumatoid arthritis

RP

relapsing polychondritis

RPDAI

relapsing polychondritis disease activity index

SGS

subglottic stenosis

SLE

systemic lupus erythematosus

SS

Sjögren’s syndrome

sTREM-1

soluble form of triggering receptor expressed on myeloid cells-1

Th1 cells

Help T cells 1

Treg cells

regulatory T cells

TM

tracheomalacia

UBA1

ubiquitin activating enzyme 1

uTIINE

urine type II collagen neoepitope

WG

Wegener’s granulomatosis

Ethics approval

Not applicable.

Conflict of interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

This work was supported by the National Key Research and Development Program of China Grants (2018YFE0207300), the National Natural Science Foundation of China Grants (81871302), Beijing Municipal Science & Technology Commission (Z211100002521021).

Data availability

Not applicable.

Author contributions

Y.Z.L. and Y.M.L. provided direction and guidance throughout the preparation of this manuscript. X.M.L., L.L.C., H.T.Z., H.L.L., and Y.H. collected and prepared the related literature. YML drafted the manuscript. Y.Z.L., X.M.L., L.L.C., and H.T.Z. reviewed and made significant revisions to the manuscript. All authors contributed to the article and approved the submitted version.

The animal research adheres to the ARRIVE guidelines

Not applicable

Permission to reproduce (for relevant content)

Not applicable

Clinical trial registration

Not applicable