Interstitial lung disease: are we missing formes frustes of connective tissue disease?

Patients with connective tissue disease (CTD) may encompass a variety of respiratory manifestations, including: pleural disease; pulmonary hypertension; bronchial or bronchiolar disease; interstitial pneumonia; restrictive lung disease secondary to musculo-skeletal changes; and iatrogenic drug-induced lung disease. Patients also have an increased risk of opportunistic infections, often favoured by corticosteroids and/or immunosuppressive or immunomodulatory therapy 1, 2. The respective frequency of manifestations, their clinical and radiological presentations, response to treatment, and outcome are dependent on the underlying CTD.

Rheumatoid arthritis may be responsible for a broad spectrum of manifestations, including: pleural effusion; chronic interstitial pneumonia; lung rheumatoid nodules; bronchiectasis; obstructive ventilatory defect due to constrictive bronchiolitis; laryngeal involvement; and opportunistic infections, such as mycobacterial infections facilitated by anti-tumour necrosis factor-α treatments. Systemic sclerosis may be commonly associated with chronic interstitial pneumonia (especially in patients with diffuse systemic sclerosis and anti-topoisomerase-1 antibodies), or pulmonary arterial hypertension (most frequently, but not exclusively, in patients with limited systemic sclerosis with anti-centromere antibodies), with both contributing to the unfavourable prognosis of the disease 3.

Although less frequent among CTDs than rheumatoid arthritis and systemic sclerosis, the spectrum of idiopathic inflammatory myopathy (including dermatomyositis, polymyositis, and amyopathic dermatomyositis) may give rise to a variety of respiratory manifestations, which represent a common cause of death in these patients. These include interstitial lung disease (ILD), aspiration pneumonia, ventilatory failure secondary to diaphragmatic dysfunction 4, and rarely, acute respiratory distress syndrome, pleural effusion, pneumothorax, pulmonary hypertension 5, diffuse alveolar haemorrhage, and dyspnoea due to dermatopolymyositis-related cardiac disease 2, 6. In addition, some manifestations may be the consequence of treatment, such as drug-induced pneumonitis and opportunistic infections. First described in 1986 7, the high frequency of pneumomediastinum in dermatomyositis has been better appreciated more recently 8.

The association of ILD and dermatomyositis, described in 1956 9, is now recognised as a direct complication of CTD and carries a poor outcome 10. ILD occurs in 5–30% of patients with dermatomyositis 2, with prevalence depending on whether chest radiographs, lung function tests or high-resolution computed tomography of the chest are used for screening. In a recent retrospective study, diffuse parenchymal lung disease diagnosed by chest radiograph was present in 70 (7.2%) out of 973 patients 11. The presence of anti-tRNA-synthetase antibodies, including anti-Jo-1 antibodies, has been reported as a marker of a subset of patients with dermatomyositis who have an increased risk of developing ILD 12–15.

Although several histopathological patterns have been described in patients with idiopathic inflammatory myopathy-associated ILD, including organising pneumonia, usual interstitial pneumonia (UIP), organising diffuse alveolar damage and lymphocytic interstitial pneumonia 11, 15–21, nonspecific interstitial pneumonia (NSIP) is the most frequent histopathological pattern 11, 20, 22. Idiopathic NSIP has been demonstrated to have a better prognosis than idiopathic pulmonary fibrosis (IPF) 23. Due to the high frequency of NSIP, survival of patients with dermatomyositis-related ILD is also better than that of IPF 11. NSIP is also the most frequent pattern of ILD in systemic sclerosis 24 and primary Sjögren's syndrome 25, whereas UIP may be more frequent in rheumatoid arthritis 26.

Despite the fact that ILD associated with inflammatory myopathy and other CTDs has been well identified, we may still be missing some forms of CTD-related ILD while in the process of investigating the diagnosis of ILD, for several reasons. First, the ILD may precede the diagnosis of CTD, the manifestations of which being moderate or overlooked, so that the ILD may not be ascribed to CTD from the start. In some cases, the ILD may actually precede any systemic manifestation of the CTD, which may follow by several months or years, then correcting the diagnosis of what had been hitherto considered as idiopathic NSIP or IPF. This has been described in as many as 20% of patients with idiopathic inflammatory myopathy 20, 27, as well as 17% of patients with rheumatoid arthritis 26.

Secondly, CTD may be limited to “formes frustes”, which may be particularly difficult to recognise. Classification criteria for systemic sclerosis now include systemic sclerosis sine scleroderma, a condition characterised by the involvement of visceral organs (lungs, kidneys, gastrointestinal system), abnormal nailfold capillaries and antinuclear antibodies, but with minimal or no cutaneous involvement 28, 29, which may be associated with progressive ILD 30. Similarly, a subset of patients with idiopathic inflammatory myopathy may have dissociated skin and skeletal muscle involvement, and patients may have florid cutaneous manifestations of dermatomyositis for a prolonged period of time (at least 2 yrs according to the definition 31) without any evidence of underlying myositis. This condition has been termed amyopathic dermatomyositis (or dermatomyositis sine myositis), and has been included in the clinical spectrum of idiopathic inflammatory myopathy 31–33. Most patients with amyopathic dermatomyositis have no objective evidence of myopathy and do not progress to the full syndrome of dermatomyositis. Conversely, some patients may have mildly elevated muscle enzymes or abnormalities of electromyogram, magnetic resonance imaging of the muscles, or of the muscle biopsy when performed (mild infiltration with inflammatory cells, with no evidence of muscle fibre necrosis) 34, a condition sometimes referred to as hypomyopathic dermatomyositis. It has recently been shown that ILD, and especially NSIP, may be associated with and may antedate amyopathic dermatomyositis 20. Thus, particular attention should be paid to cutaneous changes in a patient with ILD, such as heliotrope rash, facial erythema and oedema, Gottron's papules, and periungual telangiectasia, especially if the histopathological pattern is NSIP 35.

Thirdly, the recognition of the underlying CTD may be even more difficult when the onset of the disease is acute or subacute, a condition which has been occasionally reported in patients with dermatomyositis 19, 21. In this regard, in the current issue of the European Respiratory Journal, Suda et al. 36 have addressed novel aspects of the clinical presentation of ILD in dermatomyositis. Hence, in a retrospective cohort of 14 patients with amyopathic dermatomyositis, they identified two distinct profiles according to the clinical onset. While five out of 12 patients had a chronic ILD with a slowly progressive deterioration (often with a histopathological pattern of NSIP when lung biopsy was performed), the majority of patients (seven out of 12) presented with a subacute or acute diffuse lung disease, with a pattern of diffuse alveolar damage on histopathology. Consolidation (associated with ground-glass opacities) was more common in patients with acute/subacute ILD. Response to treatment and outcome were much better in the chronic than in the acute/subacute form of the disease.

The limitations of the study by Suda et al. 36 include its retrospective design and small size. Although most patients had a clinical and radiological pattern that differs from IPF, only a minority underwent a lung biopsy (mostly due to the rapid onset of the lung disease). Biological investigations were not extensive despite the surprisingly low prevalence of anti-Jo-1 antibodies (myositis-associated antibodies other than anti-Jo-1 antibodies were not measured). Surprisingly, the lung disease never preceded the cutaneous manifestations, contrasting with previous observations 20, 27.

Acute respiratory deteriorations of ILD have occasionally been reported in patients with idiopathic inflammatory myopathy 11, 19, 37, as well as rheumatoid arthritis, systemic lupus erythematosus and systemic sclerosis 2, 38, and are usually associated with a poor prognosis. Acute/subacute forms of CTD-associated ILD share the features of acute exacerbations of IPF, a condition also characterised by recent onset of increased dyspnoea, diffuse bilateral ground-glass opacities superimposed on subpleural reticular and honeycombing densities, and worsening gas exchange, in the absence of an identifiable cause other than the pre-existing ILD. Diffuse alveolar damage superimposed on pre-existing UIP is the pattern found on histopathology 39–41.

Whether or not the current classification of idiopathic interstitial pneumonia 42 (mostly based on histopathology) adequately applies to ILD associated with CTD is still debatable. For example, it has been shown that the prognostic significance of the histopathological pattern of lung disease related to systemic sclerosis (NSIP versus UIP) differs from that of the idiopathic counterpart of these entities 24. Fewer fibroblastic foci were found in patients with UIP in the context of CTD than in those with IPF 43, 44, suggesting that these entities may be distinct. The study by Suda et al. 36 tends to indicate that some cases of ILD (especially with an acute onset), occurring in the context of idiopathic inflammatory myopathy or formes frustes of CTD, may include features from several of the corresponding clinicopathological idiopathic entities of the classification (namely NSIP and diffuse alveolar damage), but do not fit well into a single one of them.

Are will still missing some cases of formes frustes of CTD, and why is a precise diagnosis so important in ILD? One might argue that idiopathic and CTD-related ILD often end up with undifferentiated treatments, such as corticosteroids and immunosuppressive therapy. However, in recent years, better delineation of the diagnosis of IPF has been followed by better evidence that “conventional” treatment is ineffective in this disease, and has opened the path to prospective therapeutic trials, which may ultimately lead to effective treatments of this devastating disease. Similarly, there has been some hope that early therapeutic intervention may be beneficial in patients with (idiopathic) acute interstitial pneumonia 45. Hopefully, progress in our understanding of CTD-related ILD will also provide new treatment perspectives in the future, as recently suggested in systemic sclerosis 46.

The discovery of myositis-associated antibodies (including those specific for aminoacyl-transfer RNA synthetase, anti-signal recognition particle, and anti-Mi-2 antibodies) has led to a serological approach that is complementary to the clinical evaluation 47. The recent identification of antibodies that may be specifically associated with amyopathic dermatomyositis (C-ADM-140 antibodies, representing clinically amyopathic dermatomyositis, autoantigen of 140 kDa) might aid an earlier diagnosis of this condition, especially in acute or inaugural interstitial pneumonia 48. Similarly, nucleolar-staining antinuclear antibodies with Th/To specificity may help to differentiate patients with systemic sclerosis sine scleroderma from those with idiopathic pulmonary fibrosis 49. Thus, it is anticipated that in the near future, identification of novel biological markers may help us to better delineate idiopathic interstitial lung disease from formes frustes of connective tissue disease. As for now, careful evaluation of patients with interstitial lung disease remains the mainstay of the diagnostic process.