Systemic sclerosis increases the risks of deep vein thrombosis and pulmonary thromboembolism: a nationwide cohort study.

Original article

Systemic sclerosis increases the risks of deep vein

thrombosis and pulmonary thromboembolism:

a nationwide cohort study

Wei-Sheng Chung ^1,2 , Cheng-Li Lin ^3,4 , Fung-Chang Sung ^3,4 , Wu-Huei Hsu ² ,

Wen-Ta Yang ¹ , Chuan-Chin Lu ¹ and Chia-Hung Kao ^2,5

Introduction

SSc, scleroderma is an autoimmune CTD that varies considerably in the extent and severity of skin and visceral

organ involvement [1]. Limited cutaneous scleroderma may restrict fibrosis to the hands, arms and face. Diffuse cutaneous scleroderma is a rapidly progressive disease that may affect a large area of skin and compromise one or more internal organs [2, 3]. Endothelial dysfunction, fibroblast dysfunction and dysregulation of the immune system are the three major events thought to contribute to the pathogenesis of SSc [4].

Chronic inflammation and fibrosis in SSc alter the blood vessels, skin, skeletal muscles and internal organs. Lung involvement is a frequent complication and a leading cause of morbidity and mortality in SSc patients. The most frequent type of lung involvement is interstitial lung disease, which results in pulmonary fibrosis or nonspecific interstitial pneumonia [5, 6]. However, inflammation and scarring fibrosis can be localized in the blood

vessels. In addition, the inflammation mediators connected

with prothrombotic factors and endothelial dysfunction are involved in the development of

atherosclerosis and thrombosis [7, 8].

Deep vein thrombosis (DVT) is a condition associated with the formation of blood clots in one or more of the deep veins. Pulmonary thromboembolism (PTE) is the sudden occlusion of a pulmonary artery by a blood clot dislodged from elsewhere in the body. If not treated appropriately, PTE can cause respiratory or circulatory collapse and death. Although the major PTE risk factors are recognized, the pathology often develops in people that lack traditional risk factors. Based on a longitudinal investigation of thromboembolism aetiology, Cushman et al. [9] reported that half of PTE cases were idiopathic.

Studies have shown that chronic inflammatory diseases are associated with coagulation activation and increased DVT and PTE risk [10_12]. Because SSc is not a traditional risk factor for DVT and PTE, studies investigating the risks of DVT and PTE in patients with SSc are scant. We conducted a longitudinal nationwide cohort study in Taiwan

to investigate whether SSc increases the risks of DVT and PTE.

Patients and methods

Data sources

The National Health Insurance (NHI) programme was launched in Taiwan on 1 March 1995 and covers nearly 100% of the Taiwanese population [13]. In this study we used data from the National Health Institute Research Database (NHIRD), which has been described in previous studies [10, 11]. The disease diagnoses used in the NHIRD were coded according to the criteria of the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM).

The ACR proposed classification criteria for SSc in

1980: the major criterion was proximal scleroderma, and

the minor criteria comprised sclerodactyly; digital pitting,

scars or loss of substance on the finger pads; and bibasilar

pulmonary fibrosis. Patients that fit the major criterion

and two minor criteria can be diagnosed with SSc [14]. An

SSc diagnosis was defined according to the ACR diagnostic criteria and was confirmed through the Registry for Catastrophic Illness Patient Database (RCIPD). In Taiwan, rheumatologists can apply for a catastrophic illness card for patients who fit the ACR diagnostic criteria of SSc and applications for a catastrophic illness card should be scrutinized by peer review. To ensure vulnerable populations

have access to necessary care, the NHI programme exempts beneficiaries from obligations for NHI-defined catastrophic illnesses. SSc patients who carry catastrophic

illness cards are exempted from co-payments. The NHRID encrypts the patients’ personal information for privacy protection and provides researchers with anonymous identification numbers associated with the relevant

claim information, which includes the patient’s sex, date of birth, registry of medical services and medication prescriptions. Patient consent is not required to access the NHIRD. This study was approved by the Institutional Review Board of the China Medical University and Hospital (CMU-REC-101-012).

Participants

We conducted a retrospective cohort study to investigate the association between SSc and PTE. Patients with newly diagnosed SSc (ICD-9-CM 710.1), identified using the NHIRD and confirmed in the RCIPD during 1998_2010, were selected as the SSc cohort. The index date for SSc patients was the SSc diagnosis date. We identified patients with a diagnosis of DVT (ICD-9 453.8) or non-iatrogenic PTE (ICD-9 415.1, excluding ICD-9 415.11) to

evaluate the association among DVT and non-iatrogenic PTE and SSc. Patients with a history of DVT or PTE diagnosed before the index date or missing information for age

or sex were excluded. For each patient in our SSc cohort, four patients (control patients) were randomly selected

from insurants who were not diagnosed with SSc and frequency matched based on age (every 5-year spans), sex

and the year of the index date. Control participants with a

history of DVT or PTE or missing information for age or sex

were also excluded.

Outcome measurement

The primary outcome was newly diagnosed DVT or PTE, obtained from hospitalization records. All patients were observed from the index date to the date of the primary outcome, withdrawal from the NHI programme or the end of 2010, whichever occurred first. Nearly all DVT and PTE patients underwent comprehensive examinations before receiving intensive care. In Taiwan, the medical reimbursements and discharge notes of patients are scrutinized

in a peer review process.

Exposure variables

In addition to SSc, demographic characteristics such as sex, age and co-morbidities were analysed. Pre-existing co-morbidities included atrial fibrillation (ICD-9-CM 427.31), hypertension (ICD-9-CM 401_405), diabetes

(ICD-9-CM 250), hyperlipidaemia (ICD-9-CM 272), cerebrovascular diseases (ICD-9-CM 430_438), heart failure

(ICD-9-CM 428), lower leg fracture or surgery (ICD-9-CM 820_823 and procedure codes 81.51, 81.52, 81.53 and

81.54), cancer (ICD-9-CM 140_208) and pulmonary hypertension (ICD-9-CM 416.0, 416.8 and 416.9).

Statistical analysis

All statistical analyses were performed using SAS software, version 9.2 (SAS Institute, Cary, NC, USA).

Kaplan_Meier curves were constructed using R computer software (R Foundation for Statistical Computing, Vienna, Austria). The levels of significance were set at a two-tailed P-value of <0.05. We described and compared the distribution of demographic factors and co-morbidities between

SSc and non-SSc patients using a chi-square test. The mean age comparison of both cohorts was analysed using a two-sample t-test. The event-free survival functions for DVT and PTE development in the SSc

and non-SSc cohorts were assessed using the Kaplan_Meier analysis and the difference between cohorts

was compared using the log-rank test. The incidence

densities of DVT and PTE in both cohorts were

calculated in the follow-up period until the end of 2010.

Person-years of follow-up were calculated for each patient until DVT or PTE was diagnosed or the follow-up

was stopped because of death or withdrawal from the insurance programme. Poisson regression analyses

were used to calculate the SSc to non-SSc cohort incidence rate ratio (IRR) with a 95% CI. Multivariate Cox

proportional hazard regression analyses were used to assess and compare the risks of developing DVT and PTE in the SSc and the non-SSc cohorts and to adjust the variables that were significantly related to SSc from the prior chi-square analyses. The hazard ratio (HR) and 95% CI were calculated in the Cox model. We also used

the Cox model to estimate the joint effects between comorbidities and SSc.

Results

The participants comprised 1895 SSc patients and 7580 non-SSc patients that exhibited similar sex and age distributions (Table 1). The average age of the SSc and non-SSc cohorts was 50.3 and 49.9 years, respectively.

Compared with the non-SSc cohort, the SSc cohort

exhibited a higher prevalence of atrial fibrillation, hypertension, diabetes, hyperlipidaemia, heart failure and

pulmonary hypertension at baseline. The Kaplan_Meier analysis demonstrated that the DVT-free rate (Fig. 1A) and PTE-free rate (Fig. 1B) were significantly lower in the SSc cohort than in the non-SSc cohort (log-rank test P<0.0001 for both).

Comparison of the incidence and HR of DVT and PTE

development between the SSc and non-SSc groups

Table 2 lists the DVT and PTE incidence densities in both

cohorts and the SSc:non-SSc IRRs and HRs of DVT and

PTE by sex, age and co-morbidity. The mean follow-up

was 5.34 years (

.

. 3.71) for the SSc cohort and 6.13

years (

.

. 3.64) for the non-SSc cohort and the overall

incidence of DVT (per 10 000 person-years) was 10.9

and 1.08, respectively. Compared with women in the

non-SSc cohort, women with SSc had a higher relative

incidence of DVT (IRR = 8.13, 95% CI 6.79, 9.73). The highest age-specific DVT incidence rate for patients >65 years of age with SSc (35.6 per 10 000 person-years). The co-morbidity-specific analyses indicated that patients without co-morbidity in the SSc cohort exhibited the highest DVT incidence rate (12.6 per 10 000 person-years).

Multivariate Cox proportional hazards regression analyses indicated that patients with SSc had a 10.5-fold higher risk of DVT compared with the non-SSc patients after controlling for age, sex and co-morbidities (95% CI 3.64, 30.3).

Sex-specific analysis showed an 8.57-fold significantly higher risk of developing DVT in the SSc cohort compared

with the non-SSc cohort in women (95% CI 2.86, 25.7). The age-specific relative risk of DVT was higher for patients

age 465 years with an adjusted HR of 26.3 (95% CI 3.17, 218.4). In patients without co-morbidity, the risk of DVT was 19.2-fold higher in the SSc cohort than in the non-SSc cohort (95% CI 5.33, 68.8).

The mean follow-up period was 5.35 years (

.

. 3.72) for the SSc cohort and 6.13 years (

.

. 3.64) for the non-SSc cohort and the overall PTE incidence was 7.20-fold higher in the SSc cohort compared with the non-SSc cohort (10.8 vs 1.51 per 10 000 person-years). The sex-specific analyses indicated that women in the SSc cohort exhibited the highest PTE incidence rate (11.5 per 10 000

person-years). Comparison of the SSc and non-SSc cohorts showed that the IRR of PTE decreased as age

increased (IRR = 7.95, 95% CI 6.66, 9.49 in patients465 years of age; IRR = 5.62, 95% CI 3.99, 7.93 in patients

>65 years of age). The IRR of PTE was 5.63-fold higher

among SSc patients who lacked co-morbidities compared

with non-SSc patients who lacked co-morbidities (95% CI

4.81, 6.63). SSc patients with co-morbidities exhibited a

12.2-fold higher IRR of PTE than did non-SSc patients

with co-morbidities (95% CI 7.40, 20.0). The SSc patients

exhibited a 7.00-fold increased risk of PTE development

compared with the non-SSc patients after adjusting for

sex, age and co-morbidities (adjusted HR= 7.00, 95% CI

2.64, 18.5). Compared with women in the non-SSc cohort, women in the SSc cohort exhibited a higher risk of developing PTE (adjusted HR= 8.19, 95% CI 2.70, 24.8). The

highest age-specific HR corresponded to SSc patients aged 465 years (adjusted HR= 7.60, 95% CI 2.53,

22.8). Patients with co-morbidities demonstrated an adjusted HR of PTE that was 12.0-fold higher in the SSc

cohort compared with the non-SSc cohort (95% CI 1.31, 110.0).

Joint effect of SSc and co-morbidities on the risk of DVT and PTE development

Table 3 shows the combined effects of SSc and co-morbidities on the risk of DVT and PTE development. We

included patients without SSc and co-morbidities as the reference group. Patients with SSc and no co-morbidities had a significantly increased risk of DVT (adjusted HR= 19.3, 95% CI 5.37, 69.4). Patients with SSc and co-morbidities experienced a significantly increased risk of PTE (adjusted HR= 12.0, 95% CI 3.27, 44.1) compared with patients that lacked SSc and co-morbidities.

Discussion

Previous research has shown that the prevalence of SSc was 3.8 per 100 000 persons in Taiwan, being lower than the 30_300 cases per 1 000 000 persons in Western countries [3, 15]. Women exhibited a higher risk of developing

SSc than did men; this finding is consistent with previous studies [2]. This was the first study that investigated whether SSc increased the risk of developing PTE among an Asian population by using a longitudinal populationbased cohort study. The study demonstrated that SSc patients had a 10.5-fold increased risk of DVT development

and a 7.00-fold increased risk of developing PTE compared

with the general population after adjusting for age, sex and

co-morbidities. Ramagopalan et al. [16] studied patients

diagnosed with various immune-mediated diseases who

were admitted to hospitals in England, indicating that SSc

patients had a 1.97-fold higher risk of developing venous

thromboembolism (VTE). However, Ramagopalan et al. [16]

did not adjust for co-morbidities to estimate the risk of VTE and used a shorter follow-up period compared with the period used in this study. Zoller et al. [12] performed a large-scale study of the relationship between PTE and autoimmune diseases in Sweden, indicating that patients with a history of SSc exhibited a 1.61-fold higher risk of PTE compared with non-SSc patients. These discrepancies may reflect true variations in disease occurrence among

distinct populations or environments [17_19]. Although the SSc patients in this study exhibited a higher

prevalence of co-morbidities and co-existent conditions associated with the development of DVT and PTE than did the comparison cohort, SSc remained an independent risk factor for developing DVT and PTE after adjusting for covariates. The chronic inflammatory response triggered by SSc may drive arterial and microvascular thrombosis [20, 21]. Inflammation increases procoagulant factors and inhibits the natural anticoagulant pathways and fibrinolytic

activity, causing thrombotic tendencies [21, 22]. In addition, chronic inflammation may cause endothelial dysfunction, resulting in the loss of the anticoagulant, anti-aggregant and vasodilatory properties of the endothelium [23, 24].

These possible mechanisms may explain why SSc patients exhibited a significantly lower DVT-free rate and lower PTEfree rate compared with the non-SSc cohort during the

follow-up period (Fig. 1) [4].

Most of the SSc patients were women, and female SSc patients exhibited a higher risk of DVT and PTE compared with non-SSc patients; this was consistent with the findings of previous studies [12]. SSc seemed to affect the

development of DVT and PTE more strongly in patients aged 465 years compared with patients aged >65 years.

As people age, their activity levels decrease and their cardiopulmonary systems deteriorate. The risk factors of

thrombosis associated with increased age, such as endothelial dysfunction and frailty, might affect the development

of DVT and PTE more compared with the

development of SSc [25_27].

The incidence of PTE increased when co-morbidities were present in both the SSc and non-SSc cohorts. The multiplicative risk of PE was significantly higher in SSc

patients with co-morbidities than it was in the patients who lacked co-morbidities. The results obtained using

multiple models to analyse the increased risk of PTE in

the SSc cohort, as well as the combined effects of comorbidities, were robust and consistent with those of previous

studies [10].

The primary strengths of these findings were the nationwide cohort longitudinal design and a follow-up period of

_10 128 person-years for the SSc cohort. In addition, the SSc diagnoses were identified using the NHIRD and confirmed by their inclusion in the RCIPD, conferring a high

level of reliability on the results. To avoid selection bias, we estimated the risk of DVT and PTE only in newly diagnosed SSc patients who had no previous history of DVT

and PTE. Because each NHI beneficiary is assigned a unique personal identification number, every patient could be traced through the NHIRD records throughout the follow-up period. Thus our findings can be generalized to the population of Taiwan.

However, several limitations should be considered when interpreting these findings. A greater proportion of the SSc patients were admitted to hospital than the comparison cohort, potentially causing an overestimation of

the risk of DVT and PTE in the SSc cohort.

Ramagopalan et al. [16] suggested that people admitted to hospitals for autoimmune disorders may have an

increased risk of subsequent VTE. However, our comparison cohort also comprised 35.7% of inpatients, potentially mediating the potential for overestimation since DVT and PTE are common complications during and after hospitalization for acute medical illness or surgery [28, 29].

Patients with SSc should be screened regularly for pulmonary arterial hypertension (PAH) with transthoracic

echocardiography [30]. Some patients with suspected

PAH may undergo CT scan to rule out PTE. Thus the patients

with SSc are more likely to be diagnosed with

non-symptomatic PTE than control patients. The lack of prescription drug data (e.g. hormone replacement therapy, the use of contraceptive drugs and glucocorticosteroid treatments) may also have influenced the risk of DVT and PTE among the patients in the SSc cohort.

In conclusion, our nationwide population-based cohort study of 1895 SSc patients indicated that SSc patients exhibit a 10.5-fold increased risk of DVT and a 7.00-fold increased risk of PTE compared with the general population.

In addition, the multiplicative risks of PTE were

significantly increased among SSc patients with co-morbidities.

Therefore, adequately treating co-morbidities is

critical for preventing DVT and PTE in SSc patients and

multidisciplinary teams should guide the assessment,

treatment and holistic care of SSc patients.