Sarcoidosis — the archetypal granulomatous disorder — while being a multisystem disease, clearly has a propensity to affect the lungs. The clinical presentation, imaging features, patterns of functional impairment and outcome of lung disease may vary significantly, leading to the not unreasonable suggestion that sarcoidosis is simply a convenient ‘umbrella’ term for what are, in fact, different disease entities. The investigations in this thesis have exploited imaging tests, and particularly high-resolution CT (HRCT), to explore the morphological aspects of lung disease in sarcoidosis.
The studies in the first section confirmed that chest radiography (CXR) ‘masks’ abnormalities that are more readily visible on HRCT – this was to be expected. However, a comparison between so-called CXR ‘stage’ and HRCT findings also revealed that a) parenchymal disease is present in a significant minority (40%) of patients with CXR stages 0/1 — that is, those with no CXR evidence of lung involvement and b) that fibrosis may be present on HRCT, albeit with varying prevalence, in all CXR stages. Further exploration uncovered a striking and independent linkage between specific morphological signs on HRCT— in particular traction bronchiectasis and honeycombing — and prognosis in sarcoidosis, a theme that is repeated in other fibrotic lung diseases.
In the final section, the aim was to gain consensus on the recognisable HRCT phenotypes in sarcoidosis. To this end, a Delphi study methodology was employed through which the opinions of a large body of international experts (thoracic radiologists & respiratory physicians) were garnered. The Delphi exercise identified seven key morphological/HRCT phenotypes divided roughly into non-fibrotic and fibrotic phenotypes. It is suggested that these phenotypes might, in the future, form the basis of a new morphological classification which more accurately reflects variations in clinical presentation, patterns of functional impairment, natural history and outcome in patients with pulmonary sarcoidosis.