We used a computerized microscopic image analysis system to directly measure the surface area of distal air spaces in methacrylate-embedded blocks randomly selected from inflation-fixed lobes that were resected from 45 patients as treatment of their peripheral lung tumors. In 28 of these patients, a preoperative computer tomography (CT) scan, at 6 and 10 cm below the sternal notch, was used to generate frequency histograms of CT numbers (measured as EMI units), a measure of lung density, in pixels from the lung or lobe that was subsequently resected. A similar CT number histogram was also derived from the lateral two fifths of the area of lobe/lung that was to be resected. The EMI unit that defined the lowest fifth percentile of this latter histogram correlated (n = 28, r = -0.77, p less than 0.001) with the mean value of the surface area of the walls of distal airspaces per unit lung volume (AWUV) in the five 1 mm x 1 mm microscopic fields with the lowest AWUV values, out of the 20 to 35 such fields examined in each patient. In the 34 of the 45 patients in whom we also measured volume-corrected diffusing capacity (DLCO/VA), this also correlated (n = 34, r = 0.84, p less than 0.001) with this value of AWUV, which measures the surface area of airspaces distal to the terminal bronchioles--reflecting an increase in airspace size, a defining characteristic of emphysema. However, a low DLCO/VA is nonspecific, whereas an abnormally low regional lung density is more likely to be specific for emphysema. In addition, highlighting those pixels of the CT display with low CT numbers (i.e., EMI units -500 [air] to -450, where zero = water) can locate areas of macroscopic emphysema, as shown by subsequent pathologic examination. Thus the quantitative CT scan can diagnose, quantitate, and locate mild to moderate emphysema, in humans, in life, noninvasively.

Background On

No abstract

In five patients with hypoxic chronic bronchitis and emphysema we measured ear O2 saturation (SaO2), chest movement, oronasal airflow, arterial and mixed venous gas tensions, and cardiac output during nine hypoxemic episodes (HE; SaO2 falls greater than 10%) in rapid-eye-movement (REM) sleep and during preceding periods of stable oxygenation in non-REM sleep. All nine HE occurred with recurrent short episodes of reduced chest movement, none with sleep apnea. The arterial PO2 (PaO2) fell by 6.0 +/- 1.9 (SD) Torr during the HE (P less than 0.01), but mean arterial PCO2 (PaCO2) rose by only 1.4 +/- 2.4 Torr (P greater than 0.4). The arteriovenous O2 content difference fell by 0.64 +/- 0.43 ml/100 ml of blood during the HE (P less than 0.05), but there was no significant change in cardiac output. Changes observed in PaO2 and PaCO2 during HE were similar to those in four normal subjects during 90 s of voluntary hypoventilation, when PaO2 fell by 12.3 +/- 5.6 Torr (P less than 0.05), but mean PaCO2 rose by only 2.8 +/- 2.1 Torr (P greater than 0.4). We suggest that the transient hypoxemia which occurs during REM sleep in patients with chronic bronchitis and emphysema could be explained by hypoventilation during REM sleep but that the importance of changes in distribution of ventilation-perfusion ratios cannot be assessed by presently available techniques.