• J B Cailes, R M du Bois, and D M Hansell.
• Department of Radiology and Interstitial Lung Disease Unit, Royal Brompton Hospital, London, England, United Kingdom.
• Acad Radiol. 1996 Sep 1;3(9):724-30.

Rationale And ObjectivesPulmonary hypertension may complicate systemic sclerosis in the absence of fibrosing alveolitis. Abnormal pulmonary perfusion is known to result in regional differences in lung density at computed tomographic (CT) scanning. We assessed possible disturbances in the normal CT density gradient in patients with systemic sclerosis and reduced gas transfer but no clinical or radiologic evidence of fibrosing alveolitis.MethodsWe analyzed the CT scans of 15 patients with systemic sclerosis. Regions of interest were drawn in the most dependent and nondependent zones in the upper and lower lobes of each lung. Average densities for the most dependent and nondependent zones were calculated for the upper and lower lobes, and a global lung density measurement was derived. Zonal and global density gradients were derived from mean density measures (mean density of dependent areas minus mean densities of nondependent areas).ResultsFive of the 15 patients had pulmonary hypertension, defined as an estimated systolic pressure in the pulmonary artery of greater than 30 mm Hg shown by Doppler echocardiography. Patients with pulmonary hypertension were characterized by higher global lung density for nondependent zones (median, -821 H versus -853 H; p = .07) but not for dependent zones (-800 H versus -815 H, p = .71). The lung density gradient was reduced globally and in the lower zones in patients with pulmonary hypertension (p = .05).ConclusionIn cases of systemic sclerosis without fibrosing alveolitis, we found a significantly smaller density gradient between dependent and nondependent lung regions in patients with pulmonary hypertension compared with patients without pulmonary hypertension. This finding suggests that the normal CT density gradient depends in part on normal compliance of the pulmonary vasculature.

32 keywords...

hide…