Systemic Sclerosis and the Lung

Introduction

In systemic sclerosis (SSc), cardiopulmonary involvement, consisting of interstitial lung disease (SSc-ILD) and pulmonary hypertension (SSc-PH) are the most frequent cause of death. Accurate management depends upon the early identi-cation of lung involvement. Apart from interstitial lung disease, the range of primary lung abnormalities in SSc includes sporadic case reports of obliterative airways disease, diffuse alveolar haemorrhage and signi cant pleural involvement. However, the possibility that these extremely rare complications result from overlap disorders cannot be excluded. In this chapter, we cover the epidemiology, presenting features, prognostic evaluation and management of SSc-ILD.

Epidemiology ofSSc-ILD

The pooled prevalence of SSc in 46 studies is 176/million, with a pooled incidence of 14/million person years [1]. SSc-­ ILD is present in approximately 50% of SSc patients, with the spectrum of disease severity ranging from sub-clinical lung involvement, detected during routine evaluation, to severe pulmonary disease progressing ultimately to respiratory failure and death. Epidemiological studies have been hampered by the rareness of SSc and also by variations in the de nition of pulmonary disease. These problems have recently been addressed by large international collaborative databases with standardization of diagnostic criteria and nomenclature [2]. In the EUSTAR database, compiled from over 5800 patients, SSc-ILD is the most frequent cause of death (35% of SSc-related deaths), followed by SSc-PH (26% of deaths). Renal disease was the most frequent cause

A.U. Wells (*) · G. A. Margaritopoulos · K. M. Antoniou

A.G. Nicholson

Interstitial Lung Disease Unit, Royal Brompton Hospital,

London, UK

e-mail: RBHILD@rbht.nhs.uk

of death in the last century but now accounts for only 4% of deaths [3]. SSc most commonly occurs in women (female:male ratio 5:1 in pooled studies) aged 30–50 years [1]. In USA series, the risk of SSc-ILD increases in association with African-American ethnicity: other risk factors for SSc-ILD, discussed later, include diffuse cutaneous SSc, older age at disease onset, shorter duration of disease and the presence of anti-topoisomerase antibodies [4–8].

In autopsy studies of patients with SSc, interstitial lung disease is present histologically in most patients [9]. All histological patterns seen in the idiopathic interstitial pneumonias have been reported in SSc-ILD [10]. However, the relative prevalence and prognostic signi cance of histological patterns differ greatly between idiopathic disease and SSc-ILD. In idiopathic interstitial pneumonia, the most prevalent histological pattern is usual interstitial pneumonia (UIP), corresponding to idiopathic pulmonary brosis (IPF). IPF has a strikingly worse outcome than other idiopathic disorders, including brotic nonspeci c interstitial pneumonia (NSIP). By contrast, NSIP is the most prevalent histological pattern at biopsy and autopsy in SSc-ILD (Fig. 12.1). In a series of 80 SSc-ILD patients undergoing a diagnostic surgical biopsy [11], cellular or brotic NSIP (Figs. 12.1a, b and 12.2a) was present in 78%, with less than 25% of patients having predominantly reversible disease. By contrast with idiopathic interstitial lung disease, mortality did not differ between NSIP and UIP but was associated with pulmonary function impairment at presentation and decline in pulmonary function variables during follow-up. These ndings are broadly compatible with outcome data in smaller series, although in one reports, the presence of UIP was associated with a trend towards a shorter survival time [12].

High-resolution computed tomography (HRCT) is the primary means of detecting SSc-ILD [13, 14]. Interstitial lung disease is present in over 90% of SSc patients with abnormal pulmonary function tests and in up to 65% of SSc patients overall. HRCT appearances are typically compatible with NSIP, with prominent ground-glass attenuation and a low prevalence of honeycomb change [15]. HRCT has a high

Fig. 12.1  Cellular NSIP. The lung shows diffuse involvement of the alveolar interstitium by a mild chronic infammatory in ltrate more evident at high power (H&E stain, ×80 magni cation)

Fig. 12.2  Fibrotic NSIP. The lung shows diffuse involvement of the alveolar interstitium by a homogenous brosis with minimal infammation. No broblastic foci were seen at high power (H&E stain, ×10 magni cation)

diagnostic sensitivity which has obvious advantages but has also caused dif culties for clinicians as very limited HRCT abnormalities are often disclosed by routine screening tests. As HRCT abnormalities are often limited and dif cult to interpret, a multidisciplinary evaluation of disease severity is essential, with the integration of symptoms, pulmonary function tests and the extent of disease on HRCT. The selection of patients for treatment, discussed later, has been facilitated

by studies of the natural history and treated course of SSc-­ ILD in large historical cohorts [16, 17] and especially by HRCT data [18, 19].

Risk factors for SSc-ILD

SSc-ILD has been linked to both the distinction between limited and diffuse cutaneous disease and the autoantibody pro-le. SSc is sub-classi ed as limited or diffuse cutaneous disease, according to the extent of skin involvement. In limited disease, skin involvement is distal to the elbows and knees (although facial involvement may occur) whereas in diffuse disease, there is variable involvement of the trunk, shoulders, pelvic girdles and the face and acral areas. Across many series, the prevalence of SSc-ILD has generally been higher in diffuse cutaneous disease, but this association is weaker than the associations between SSc-ILD and autoantibody status [20].

In the EUSTAR database, autoantibody status was a more powerful predictor of major organ involvement than the distinction between diffuse and limited cutaneous disease [21]. Antinuclear antibodies (ANA) are present in more than 90% of SSc patients. Autoantibodies against topoisomerase I (ATA, also known as anti-Scl70 antibody), present in over 20% of SSc patients, are associated with the development of pulmonary brosis in over 85% of cases [22]. However, only 40% of patients with SSc-ILD are ATA positive [23]. Reported correlations between ATA levels and the severity of SSc-ILD [24, 25] are not suf ciently consistent to infuence the investigation of individual patients. Anti-centromere antibody (ACA) positivity, associated with a very low prevalence of signi cant SSc-ILD, is linked to limited cutaneous disease and an increased risk of SSc-PH [4, 26].

Genetic Associations

The accumulated evidence indicative of a genetic predisposition to SSc includes associations between speci c autoantibodies and major histocompatibility complex (MHC) classes, clustering of SSc with other rheumatic diseases in family members and familial cases of SSc, including in twins [27]. The Choctaw Indians have a tenfold increase in prevalence of SSc and a genome-wide screen has disclosed multiple microsatellite markers in chromosome regions associated with SSc, including the MHC, brillin 1 gene (15q), the topoisomerase 1 gene (chromosome 20q) and the SPARC gene (secreted protein, acid rich in cysteine; chromosome 5q) [28]. However, the overall familial linkage to SSc is weak, with less than 3% of patient siblings developing the disease, and the current consensus is that many gene variants