Clinical characteristics and disease patterns of patients with relapsing polychondritis: a retrospective cohort


 Background Relapsing polychondritis (RP) is a rare autoimmune disease affected various cartilage, Patients with tracheal cartilage involvement are different from other patients. The objectives of this study were to allocated RP patients into two subgroups by chest computed tomography (CT) and compare the clinical features and disease patterns of each group.Methods A retrospective cohort study collected RP patients hospitalized at the Beijing Chao-Yang Hospital between January 2012 - August 2021. Patients were divided into two groups: respiratory involvement group and non-respiratory involvement group according to chest CT.Results In our study, respiratory involvement found in 59.7% (n=43) patients, which had higher rate of costochondritis, fewer rate of Inflammatory eye disease and auricular chondritis than those in non-respiratory involvement. Compared with non-respiratory involvement subgroup, The incidence of pulmonary infection marginally increased and those inflammatory indexes except for CAR were significantly higher in respiratory involvement subgroup, further subgroup analysis found that there was no significant relationship between inflammatory indexes and pulmonary infection. Finally, 5 patients died during the follow-up in this cohort with a median follow-up time of 6 years (range 3-8 years).Conclusion 59.7% of patients had respiratory involvement according to chest CT findings in our cohort, which had a strong inverse relationship between respiratory and auricular, ocular involvement. Increase inflammatory indexes were not correlated with pulmonary infection, suggesting that patients with respiratory involvement had a higher disease activity index of RP. The probability of survival was not found significant in two subgroups.


Introduction
Relapsing polychondritis (RP) is a rare, systemic immune-mediated disease characterized by recurrent and progressive in ammation of the hyaline, elastic and brous cartilaginous structures, predominantly the external ear, nose, and tracheobronchial tree (1).Skin, cardiovascular system, nervous system, blood system may be affect. The rst case of RP was reported by Jaksch-Wartenhorst in 1923, and Pearson et al introduced the term "RP" and summarized the common clinical manifestations in 1960 (2). In a UK study, the incidence of RP between 1990 and 2012 was 0.71 per million population per year with a standardized mortality ratio of 2. 16 (3) and the prevalence of RP in the Department of Defense bene ciary population was 4.5 per million(4).
Respiratory involvement was present in up to 50 % of RP patients (5,6). Symptoms such as cough, stridor, progressive dyspnea and hoarseness were common in these patients, who were misdiagnosed as chronic bronchitis and refractory to conventional treatment, leading to a diagnostic dilemma and poor prognosis.
Chest CT could reveal the classic morphologic respiratory changes associated with RP. Thus, the purpose of the present study was to retrospectively divide RP patients into 2 subgroups by chest CT ndings, namely a subgroup of patients with respiratory involvement and a subgroup of patients with nonrespiratory involvement, then describe and compare the clinical features and disease patterns of each subgroup.

Patient population
We retrospectively recruited RP patients who were admitted to the Beijing Chao-Yang Hospital from January 2012 to August 2021. RP was diagnosed according to the diagnostic criteria proposed by McAdam et al. (7), Damiani and Levine(8) and Michet et al.(9) (Table 1). Informed consent can be waived for this retrospective study with the approval of the the ethics committee.

Data acquisition
Data were obtained from the Electronic Medical Records (EMR). Patients younger than 18 years or without complete electronic case les were excluded. Patients with other connective tissue disease were also excluded. The clinical data were collected for all patients from medical records: patients' pro les, clinical features, chest CT scan, blood test, autoimmune series including rheumatoid factor (RF), antinuclear antibodies (ANA), anti-dsDNA antibodies, antineutrophil cytoplasmic antibodies (ANCA) and therapeutic interventions.

Subgroup de nition
All patients performed chest computed tomography and respiratory involvement in our study included respiratory wall thickness or calci cation and respiratory stenosis due to the lack of end-inspiratory and dynamic expiratory scans. Respiratory wall thickness was de ned as the thickness of the wall of the involved respiratory segment was greater than 2 mm, respiratory narrowing was de ned as at least a 25% reduction in the diameter of the lumen(10).

Statistical analysis
The descriptive analysis included the absolute and relative (percentage) frequencies for the categorical variables, the means (standard deviations, SD) and medians (interquartile ranges, IQR) for the quantitative parametric. Differences between the groups were computed using the Student's t test or Fisher's exact test for quantitative variables and the chi-square test for the qualitative variables and the Mann-Whitney or Kruskal-Wallis tests for the non-normal variables. Spearman correlation coe cient was used to describe the correlation between two categorical variables. Kaplan-Meier (K-M) survival analyses and Log-rank test were constructed to reveal survival analysis. Multiple imputation was performed on missing data. All analyses were performed using SPSS software version 21.0 and GraphPad Prism version 7.0 Two-tailed, p-value < 0.05 was considered statistically signi cant.

Results
From January 2012 to August 2021, 75 patients with RP were screened in this study. 2 patients with positive for ANCA against MPO (myeloperoxidase) diagnosed as Systemic vasculitis and 1 patient with positive for ANCA against PR3 (proteinase-3) diagnosed as ANCA associated vasculitis were excluded. The diagnosis based on Damiani and Levine's criteria was con rmed in 45 patients (62.5%), either Michet or Damiani criteria were ful lled in 66 patients (91.6%). 16 patients (22.2%) met all the three criteria, 6 (8.3%) patients did not ful l any set of criteria, but the improvement of auricular in ammation and respiratory symptoms after corticosteroid treatment was typical for RP. We reviewed previous chest CT images and found respiratory wall thickening sparing the posterior membranous wall was identi ed in 43 patients (59.7%) ( Figure.

Demographic characteristics
A total of 42 male and 30 female patients were included in our study, 29 of 72 patients smoked. The average age at the time of rst symptoms was 54.03±13.10 years and the average age at diagnosis was 55.20±12.59 years. The delay from the time of the rst symptom to the diagnosis was 6(2-24) months, and the duration of follow-up since the establishment of diagnosis was 4 (3-6) years. Finally, 5 patients died during the follow-up. We did not nd signi cant differences in the demographic features of two subgroups ( Table 2).

Subgroup analyses of clinical characteristic and correlation analysis
Respiratory involvement (n=43) and auricular chondritis (n=26) were the most frequent manifestations, followed by ocular involvement consisting of scleritis and uveitis (n=18), sensorineural hear loss (n=18), costochondritis (n=14), nasal chondritis (n=11). Patients in respiratory involvement subgroup had a signi cantly higher occurrence of costochondritis (p=0.03) and a less occurrence of auricular chondritis (p=0.001), In ammatory eye disease (p=0.001) than non-respiratory involvement subgroup. Laryngeal involvement in 8 patients, In ammatory arthritis in 6 patients, Cutaneous manifestations in 4 patients were other common seen manifestations, there was no signi cant difference in above clinical manifestations between the two groups. Although not statistically signi cant, there was a clear trend toward a higher frequency of pulmonary infection in respiratory involvement subgroups (p=0.06). Cardiac involvement in the form of myocardial infarction or ventricular tachycardia was seen in 4 patients. One patient developed cytopenia after treatment with glucocorticoids and immunosuppressants, Bone marrow biopsy showed bone marrow suppression.
We performed correlation analysis between different organ involvement and found a negative correlation between respiratory involvement and auricular chondritis (r=-0.58, p < 0.01), and also between respiratory involvement and in ammatory eye disease (r=-0.45, P < 0.01). Auricular chondritis was positively correlated with in ammatory eye disease (r=0.49, P < 0.01) A weak positive correlation was also revealed between auricular chondritis and pulmonary infection (r=-0.39, P < 0.05). (Figure.2)

Comparisons of laboratory ndings and subgroup analysis
The respiratory involvement subgroup had higher CRP and ESR concentrations than non-respiratory involvement subgroup (p=0.03, p=0.04). Creatinine and Uric acid in the non-respiratory involvement subgroup were signi cantly higher than in respiratory involvement subgroup (p=0.02, p=0.01). Novel in ammatory markers associated with RP disease activity index NLR, PLR were higher in respiratory involvement subgroup (p=0.03, p=0.01), but CAR was lower (p=0.04, p=0.01). We further conducted subgroup analysis on whether patients with respiratory involvement had pulmonary infection, and found no statistical difference in in ammatory indicators between the two groups of patients. (Table.3) PFTs were performed in 22 patients with respiratory involvement, and showed respiratory obstruction in all patients, obviously reducing in forced expiratory volume in 1 second (FEV1) 1.21±0.54L, forced expiratory volume in 1 second/forced vital capacity (FEV1/FVC) 42.11±14.84%, Residual volume (RV), total lung capacity (TLC), Residual volume /total lung capacity (RV/TLC) were usually normal.
Positive antinuclear antibodies were found in 8 patients in a titre of 1: 320, rheumatoid factor was positive in 4 patients, antineutrophil cytoplasmic antibodies were positive in 4 patients, but no particular speci cities were found.

Survivals
After a follow-up period of 6 (3-8) years since the rst symptoms, 6.9% of the patients (n = 5) had died (Table.2), K-M curve and Log-rank test also showed the probability of survival was not statistically different between patients with and without respiratory involvement ( Figure.3).

Discussion
In the present study, we compared the clinical features and disease patterns of RP patients with or without respiratory involvement. We found that RP is equally prevalent in men and women, with onset occurring between the ages of 40-60, which is consistent with other series (4, 11). Age at the time of rst symptoms, the time of the rst symptom to the diagnosis and age at diagnosis were also no signi cant differences between two groups. Respiratory involvement subgroup was characterized by higher rate of costochondritis, fewer In ammatory eye disease and auricular chondritis compared to non-respiratory involvement. Respiratory involvement subgroup trended towards signi cantly increasing the likelihood of pulmonary infection. CRP, ESR and novel in ammatory markers associated with RP activity index (NLR, PLR) were higher in respiratory involvement subgroup, but CAR was Lower. The level of Creatinine and Uric acid were higher in non-respiratory involvement subgroup. PFTs suggested that patients with respiratory involvement had obvious ventilation dysfunction, and gas exchange was not involved. 64 patients received treatments, with a base of prednisone of 0.5-1mg/Kg/d, 25%(n=18), 19.4%(n=14) of patients were being treated with MTX and CYC, respectively. The median follow-up period was 6 years (ranger 3-8 years), 5 of the patients had died, the probability of survival was not statistically different between two subgroups.
RP is a rare autoimmune disease characterized by recurrent in ammation and destruction of cartilage (12). The cause of RP is uncertain but regarded as autoimmune-mediated (13). Although anti-Collagen Type II antibodies, matrilin-1, cartilage oligomeric matrix proteins could be found in the targeted cartilage, the above antibodies are not speci c and not used in clinical practice (13)(14)(15)(16)(17). There is no validated classi cation for RP, the diagnosis depends on a combination of clinical representation, imaging and biopsy of involved cartilaginous tissues. Ernst (18) investigated 145 RP patients and found 21% of patients had respiratory involvement, but respiratory symptoms were the rst symptom in 54% of these patients. Dion (19) performed a retrospective study of 142 RP patients using a cluster analysis in France to identi ed three separate phenotypes: hematological phenotype, respiratory phenotype, mild phenotype and found that 22% of RP patients exhibited respiratory manifestations and thought that patients in the respiratory phenotype frequently received intensive treatments and suffered from more infections, which had been considered as a signi cant cause of mortality in previous studies. Recently, one study (20) included 239 RP patients found 19.7% and 29.3% of these patients in the respiratory involvement without auricular involvement subgroup and both respiratory and auricular involvement subgroup, respectively. The above studies de ned respiratory involvement by the patterns of clinical manifestations. Meanwhile, a good agreement was found between clinical and radiographic data for the diagnosis and assessment of disease activity in RP patients (21). A review(10) suggested that chest CT scan particularly with end-inspiratory and dynamic expiratory scans should be routinely performed at diagnosis and during the evolution of the disease. Malacia or air trapping may be the unique CT abnormalities (22,23), which could be more easily detected during dynamic expiration scan. In our study, 43 patients were found thickening of anterior and lateral respiratory walls, which was considered to be the early stage of the disease (17), increased attenuation of the cartilages sparing the posterior membranous wall and respiratory stenosis was identi ed in 24 patients,10 patients.
Consistent with studies in the French and Japanese patients (19,20), our current study demonstrated there were obvious differences between two subgroups in terms of clinical characteristics: Respiratory involvement subgroup was characterized by laryngotracheal chondritis and costochondritis, nonrespiratory involvement subgroup was distinguished by in ammatory eye disease and auricular chondritis. We also found ve patients with auricular chondritis or in ammatory eye disease were assigned to the respiratory involvement subgroup, due to their chest CT showed respiratory wall thickening, calci cation or stenosis. We didn't do a further subgroup analysis on account of small quantity, those patients may require to be isolated and analyzed as well as the Overlap subgroup (20). In contrast to other studies(4, 7, 24-26), involvements of skin (5.6 %: 13-46 %) and cardiovascular involvements (5.6 %: 2-31 %) were less common. Chang-Miller et al (27) reported renal involvement was observed in 29 of 129 patients, however, renal involvement were not present in our study, but uric acid and creatinine levels were higher in the non-respiratory involvement subgroup, reminding possible kidney damage.
Although RP is considered to be an immune system disease, no signi cant abnormalities have been found in either cellular or humoral immunity. CRP and ESR are markers of the severity of common rheumatic diseases, Higher CRP and ESR concentrations in the respiratory involvement subgroup suggested more in ammation and disease activity. Novel in ammatory markers were associated with activity in many rheumatic diseases (28,29). Cao(30) indicated that CAR, NLR and PLR levels were signi cantly higher in RP patients than healthy controls and were positively correlated with RPDAI. We found that NLR, PLR were higher in respiratory involvement subgroup, but CAR was lower compared to non-respiratory involvement subgroup. Subgroup analysis of patients with respiratory involvement found that there was no statistical difference between pulmonary infection and in ammation indexes suggesting those patients had a high level of disease activity.Our study had some limitations. Firstly, it was a single-center, retrospective study, with a small sample size. Secondly, Air trapping or Mosaic sign occurred in the early stages of RP. In our study, when RP was diagnosed without dynamic expiration, Air trapping or Mosaic sign cannot be detected in time and may result in a delay in diagnosis In Conclusion, 59.7% of patients had respiratory involvement according to chest CT ndings in our cohort, which had a strong inverse relationship between respiratory and auricular, ocular involvement. Increase in ammatory indexes were not correlated with pulmonary infection, suggesting that patients with respiratory involvement had a higher disease activity index of RP. The probability of survival was not found signi cant in two subgroups. AUTHOR CONTRIBUTIONS DW and LJG participated in the study design, performance, and manuscript writing. XFZ conducted medical data collection, XD participated in patient screening as a rheumatologist, ZHT revised the manuscript. All authors read and approved the nal manuscript.

Funding
There is no funding allocated for this retrospective study.

Availability of data and materials
Raw data is available from the corresponding author on reasonable request Ethics approval and consent to participate Informed consent can be waived for this retrospective study with the approval of the the medical Ethics Committee of the Chaoyang Hospital, Beijing.

Consent for publication
Not applicable.

Competing interests
None of the investigators declare any real or perceived con icts of interest pertaining to the subject of this manuscript.  Tables   Table 1 Three    Kaplan-Meier survival curve and Log-rank test of respiratory involvement and non-respiratory involvement RP patients