Abstract 1725

Glucocorticoids are widely used in the perinatal period to advance pulmonary maturation and improve alveolar function. These processes imply an orderly formation of extracellular matrix, particularly the basement membrane structures at the air-blood interface. However, glucocorticoids rank among the most potent inhibitors of extracellular matrix formation. A suppressive effect of glucocorticoids on the main structural component of basement membranes, collagen type IV, should conflict with its beneficial effect on lung maturation. We therefore hypothesized that glucocorticoids when administered to neonates, do not negatively impact collagen type IV. In an exploratory investigation of this prediction, we enrolled 10 low-birthweight infants (8 females, 2 males) admitted to our intensive care unit, into an IRB-approved protocol after obtaining parental informed consent, and for a total of 536 days, monitored the levels of the C-IV antigen in 81 consecutively obtained serum samples. C-IV is derived from collagen type IV and, as evidenced by its co-fluctuation with nocturnal growth hormone peaks in children, at least in part reflects biosynthetic events in vivo. Of the 10 pats., 5 required glucocorticoids and underwent a total of 8 treatment periods comprising 75 days. The major clinical determinants between the steroid-medicated group (SMG) and the non-medicated group (NMG) did not differ (gestational age, 24.8±1.4 vs. 26±1.6 weeks; birth weight, 775±160 vs. 729 ±134 grams; mechanical ventilation, 33.2±17.7 vs. 20.6±16.3 days). Differences between the mean C-IV levels also did not achieve statistical significance. To assess the kinetic behavior of the C-IV antigen in both groups, we analyzed for each patient the slope of this parameter between two consecutive measurements and from this value, calculated the change of C-IV slope per day. For this parameter, four distinct types were defined and their medians calculated: 'from no treatment into treatment' (INTO); 'within treatment' (IN); 'from treatment into no treatment' (OUT) and 'no treatment' (NO). The medians of the NO type of C-IV slope per day were small and similar for SMG patients off steroids and the NMG group (-0.6 vs. 1.9). By contrast, the median of the INTO type of C-IV slope was -3.15, indicating an initial decrease in the serum levels upon start of steroid administration. Suprisingly, this decline appeared temporary and reversed even while steroids were continued, causing the IN type of C-IV slope to invert to a median of +4.15 during treatment. Even more remarkable, once steroid treatment was discontinued, the C-IV levels displayed conspicious increases in excess of pre-treatment levels, reflected by the rise of the OUT type of C-IV slope to a median of +18.2 (significance by Kruskal-Wallis, p =0.014). This, glucocorticoids do not consistently depress C-IV, but trigger a notable C-IV rebound in neonates.

Supported by NIH RR06020