Characteristics of connective tissue diseases associated interstitial lung diseases – a single centre study in Sri Lanka

Introduction Interstitial lung diseases are the major cause of mortality in connective tissue diseases. Objectives Our aim is to describe patients with interstitial lung disease associated with connective tissue diseases (CTD-ILD) in the Chest Clinic, Colombo 08. Methods We conducted a descriptive cross-sectional study at the Central Chest Clinic, Colombo, Sri Lanka and data of all patients attending the Clinic were analysed. Results The sample consisted of 83 consecutive patients diagnosed with CTD-ILD. The age ranged from 24 to 72 years with mean (SD) age of 55.6 (10.6) years. The majority was female patients (n=70, 88.6%). The majority 53.0% (n=44) had RA and SSc was the second commonest CTD-ILD (n=22, 26.5%). NSIP 51(61.9%) was the most frequently observed HRCT pattern in our cohort whereas UIP was the second commonest. RA-ILD subgroup demonstrated female (n=37, 86.0%) preponderance. NSIP (n=24, 55.8%) was the commonly observed HRCT pattern followed by UIP (n=15, 34.9%) pattern. Among RA-ILD patients, 28 (65.1%) had positive Rheumatoid factor. Majority of (n=20, 90.9%) SSc-ILD were females. Sixteen (72.7%) of them had NSIP pattern in HRCT analysis followed by UIP (n=6, 27.3%). Conclusion RA-ILD was the most frequent type of CTD-ILD followed by SSc. More importantly; the present study revealed the predominant NSIP pattern and clear female preponderance in RA-ILD compared to global data. We recommend prospective multicenter studies to be carried out and prospective disease registries to be established to explore the epidemiological, clinical, radiological and prognostic characteristics of CTD-ILD in Sri Lanka.


Introduction
The connective tissue diseases (CTDs) are multisystemic rheumatological diseases characterized by autoimmune mediated disease process causing multiorgan dysfunction. Among the multiple causes leading to mortality in CTD patients, interstitial lung diseases (ILD) are identified as the main cause [1].
Identification of the type of ILD is necessary to decide on definitive treatment and more importantly to prognosticate the disease. Interestingly, most of the time High Resolution CT scan (HRCT) patterns reflect the underlying histopathological subtypes [4]. Studies have shown that in UIP, HRCT diagnosis has excellent concordance with the histological diagnosis [5]. Lung biopsy is recommended only in atypical cases [6]. Thus, identification of the pattern of HRCT is mandatory to decide definitive treatment and assess the prognosis.
It is Important to note that limited studies have been conducted regarding histological and radiological pattern of CTD-ILD in Sri Lanka [7]. Available data also being obtained from the studies conducted in the other countries [8]. In practice, these findings are extrapolated to various clinical settings across the globe including Sri Lanka. However, it is important to generate local evidence base in the Sri Lankan population in order to identify any possible differences compared to the western population.

Original article
Many sociodemographic and immunological factors are identified as risk factors of developing ILD [8]. As an example, male gender, severity of the RA, older age, presence of rheumatoid factor, anti CCP antibody and smoking carry risk for developing RA-ILD [9]. This further emphasizes the necessity of data from our population, which has many sociodemographic differences compared to European countries. Hence, it is important to identify which subtype of ILD is predominant in the Sri Lankan population. Accordingly, the available diagnostic and treatment facilities could be further developed to cater the needs of the local setting.

Methods
We conducted a descriptive cross-sectional study at Central Chest Clinic, Colombo, Sri Lanka, which is the largest chest clinic in the country. CTD-ILD patients, who met 2010 American College of Rheumatology and European League Against Rheumatism (ACR/EULAR) classification criteria for RA, 2013 ACR/EULAR classification criteria for SSc, 2017 ACR/EULAR classification criteria for DM/ PM, 2016 ACR/EULAR classification criteria for pSS and 2015 ERS diagnostic criteria for IPAF were included in the study. Diagnosis of ILD was made on the basis of the HRCT findings.
Information on socio-demographic data, clinical features and autoantibodies were recorded by reviewing patient medical records. All HRCT findings, which were evaluated at the multidisciplinary team meeting by an experienced Radiologist and a Consultant Chest Physician were documented.
Data was collected after obtaining the ethical clearance from Medical Research Institute (MRI) in Sri Lanka, informed written consent of the patient and permission from the hospital director.
Data analysis was carried out using SPSS version 25.0. Descriptive statistics including mean [Standard Deviation (SD)], median [Inter-quartile range (IQR)] and frequency distributions were used to present study variables. Fisher's exact test was used to compare the differences of categorical variables between subgroups.

Sample characteristics
The sample consisted of 83 consecutive patients diagnosed with CTD-ILD attending the Central Chest Clinic Colombo 08. The age ranged from 24 to 72 years and the mean (SD) age was 55.6 (10.6) years. Mean age of female and male were 55.0 and 58.7 respectively. The majority was female patients (n=70, 88.6%).
Time of the initial diagnosis was recorded in 55 subjects. CTD was diagnosed before ILD in 53 (96.4%) of them, with a median (IQR) interval of 4 (2-8) years separating the two diagnoses. ILD was diagnosed first only in two (3.6%) patients.

Distribution of CTD-ILD
The distribution of the CTD-ILD in the sample is given in Table 1. The majority 53.0% (n=44) had RA and SSc was the second commonest CTD-ILD (n=22,26.5%).

HRCT patterns of ILD
NSIP (n=51, 61.9%) was the most frequently observed HRCT pattern among the CTD-ILD patients, whereas UIP was the second commonest pattern (n=25, 30.8%) Details of the HRCT patterns are given in Table 2.

Clinical features
Shortness of breath and cough were the commonest presenting symptoms found in 55 (66.3%) and 49 (59.0%) patients respectively. Fine crepitations on auscultation of lungs was noted among 43 (51.8%) of the sample. One fifth of the sample had fatigue (n=17, 20.5%).
There were no statistically significant associations between RA and gender, HRCT patterns or symptomatology (p>0.05). However, there was a statistically significant association between RA and rheumatoid factor positivity (Chi (1) = 24.8675 p<0.001).

Discussion
CTD-ILDs were responsible for vast majority of secondary ILD in Sri Lanka [7]. This is the first study to describe the characteristics of CTD-ILD patients attending to lung specialized tertiary care centre in Sri Lanka.
The mean age of 55.6 years of our sample was similar to available literature [8,10,11,12]. Female preponderance was demonstrated as expected since the CTD was well known to occur in female gender [7,8,11,12].
Evidence suggests that in some instances, ILD is the initial clinical manifestation of CTD [10,12]. There were two patients who initially presented with ILD in our sample as well. However, the date of the initial diagnoses was recorded only in 55 patients. RA was the most prevalent subtype of CTD-ILD followed by SSc. RA-ILD was recognized as the commonest type in many studies conducted globally [12,13]. A recent study conducted in Sri Lanka reported similar findings [7]. However, a descriptive study conducted in Portugal reported SSc as the commonest CTD-ILD subtype [8].
In consistent with the existing literature, our study reveals that the predominant pattern of HRCT in CTD-ILD group was NSIP followed by UIP [8,12]. As expected, shortness of breath and cough were the commonest presenting symptoms found in patients [11]. At the same time, it is important to note significant proportion did not have expected symptoms, which signifies the importance of screening with LFT and HRCT even in asymptomatic group.
Pulmonary function test which is widely used as a screening tool in detecting ILD, revealed predominant restrictive pattern as reported in literature [7,8].

Original article
RA-ILD is an important entity that carries higher mortality than RA patients without ILD [14]. RA-ILD patients, who comprised the majority of our sample, reported clear female preponderance. Despite female predominance in RA, the largest RA-ILD study conducted in the UK has revealed almost equal gender distribution [9]. On the other hand, there were some studies showing female preponderance as well [14,15] emphasizing the importance of conducting local research to identify epidemiological variations.
Globally, many studies demonstrated UIP pattern as the predominant HRCT pattern in RA-ILD [2,9,14,16,17,18]. In contrast, the majority of our patients had NSIP pattern in HRCT followed by UIP in approximately one third of patients. Recent study conducted in Sri Lanka also reported NSIP pattern as the commonest HRCT pattern closely followed by UIP [7]. There were few cases of drug induced HP and organizing pneumonia. Though it is rare [19] we reported one case with CPFE Combined pulmonary fibrosis and emphysema as well.
RA-UIP pattern carries poorer prognosis and increases all-cause mortality in RA patients [9]. Since the NSIP pattern is the commonest type, our RA-ILD patients may have better overall survival in contrast to other countries, which we have not addressed in our study.
Our study demonstrated statistically significant difference in the presenting age of RA-ILD and other CTD-ILD. RA-ILD presented in late fifties compared to other CTD-ILDs, which presented in early fifties.
It is an already known fact that high levels of Rheumatoid Factor (RF) and titers of anti-CCP increase the risk of ILD development in RA [8,9,20,21]. Confirming this, the majority of our RA-ILD patients had positive RF and that association was statistically significant. We have not analyzed Anti-CCP as it was available only in few patients, since it is not freely available in the government sector.
SSc-ILD is the major cause of mortality in SSc [22,23]. The mean age of the SSc-ILD was significantly lower than that of other CTD-ILD groups. This showed the early presentation of SSc ILD compared to other group. Female preponderance was clearly observed in our sample similar to other studies [2,14,24]. The majority of SSc-ILD group was symptomatic at the time of presentation. Similar to global evidence NSIP was the predominant HRCT pattern [2,14,24].

Limitations
The present study was a single-center retrospective study; hence, generalizing the findings of the study needs to be done cautiously. Since evidence suggest that there are substantial differences in data obtained from prospective disease registries compared to single-center retrospective studies [12], we believe it is important to establish prospective disease registries and conduct future studies to explore the potential associations further. Since this study involved retrospective data, some of the relevant data could not be retrieved due to misplacement of reports and records as well as deficiencies in record keeping.

Conclusions
To the best of our knowledge this is the first study to describe CTD-ILD patients in Sri Lanka. Comparison of the present study findings with evidence from other countries revealed both similarities and differences. RA-ILD was the most frequent type of CTD-ILD observed in our center followed by SSc. The most striking finding is that the predominant NSIP pattern and clear female preponderance in RA-ILD compared to global data.
We recommend prospective multicenter studies to be carried out and prospective disease registries to be established to explore the epidemiological, clinical, radiological and prognostic characteristics of CTD-ILD in Sri Lanka.

Declarations
Ethics approval and consent to participate: Ethicalapproval was obtained from the Ethics Review Committee of the Medical Research Institute (MRI) in Sri Lanka. Informed written consent was taken from the participants prior to entering in the database.

Consent for publication: Not applicable.
Availability of data and materials: The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.

Conflicts of interest:
All authors declare that there are no conflicts of interest.