Objective: Infants with chronic lung disease (CLD) have an arrest of primary and secondary septation. We hypothesized that this may be related to damage or abnormal development of lung collagen secondary to positive pressure ventilation. Our aims were to identify the sites and quantity of collagen in control infants 22 to 72 weeks' postconceptional age and compare these with infants with various degrees of severity of CLD.
Methods: The controls were 22 to 42 weeks' gestation (n = 30), received minimal ventilator care, and died within 48 hours of birth, plus 5 term infants who died at 43 to 72 weeks' postconceptional age from nonpulmonary causes. Infants who were 23 to 30 weeks' gestation, were at risk for CLD, and lived 5 to 94 days (n = 33) were separated into 3 groups on the basis of respiratory score (score group; the integrated area under the curve of the average daily fraction of inspired oxygen x mean airway pressure [cm H2O] over the number of days lived). The score groups, <20, 20 to 69, and 70 to 500, related clinically to mild to moderate and severe lung disease. The lungs were tracheally perfused and formalin fixed. Total lung volume was determined by water displacement. The paraffin-embedded lung blocks were sectioned 5 micro m thick, stained with Gomori's reticulum stain, hematoxylin and eosin, and immunohistochemically for collagen IV. The parenchyma was point-counted, and the volume density of collagen was measured. The chord diameter of the peripheral airway saccules and alveoli was measured. Descriptive collagen data were assessed on en face 40- micro m-thick sections through the alveolar or saccular walls on all infants at risk for CLD and in selected controls.
Results: In the controls, the volume density of collagen decreased from a maximum of 9% at 22 weeks to 5% at term and 72 weeks. With Scores < or =69, the fraction of collagen was similar to controls, but in infants with scores 70 to 500, it was increased relative to controls. However, when collagen was expressed as the volume density of interstitial tissue, ie, excluding parenchymal air space, it increased from a low of 5% at 22 weeks to 25% at 72 weeks. In infants with scores 70 to 500, 79% of infants had collagens greater than controls. Saccular and alveolar diameter increased from 40 micro m at 23 weeks to 100 microm at 72 weeks. Most infants with severe CLD (scores > or =70) had diameters more than twice that of controls at the same age. The total lung parenchymal collagen had a similar pattern as the volume density of collagen in interstitial tissue, increasing from 0.4 cm3 at 23 weeks to 9.7 cm3 at 72 weeks in the controls. Eighty-five percent of infants with scores 70 to 500 had total parenchymal collagen greater than the 95% confidence interval of the controls. With en face sections, a fine collagen mesh was seen at 23 weeks, which progressively increased in fiber size and quantity until 72 weeks. With severe CLD, the secondary collagen fibers in the saccular wall were thickened, tortuous, and disorganized relative to same-aged controls. Under 30 weeks, in the controls, the interstitium contained a wide, delicate network of interconnected collagen fibers. After positive pressure ventilation, some saccules markedly increased their diameter, which compressed and obliterated the interstitial network. In contrast with severe CLD, the interstitium was wide, with coarse wavy collagen fibers.
Conclusions: Parenchymal collagen increases throughout development. Before 30 weeks, there is a delicate complex interstitial collagen network, which may be important for primary septation and subsequent normal development. Positive pressure ventilation, if excessive, and depending on lung maturity and disease state, over a short time can severely compress the interstitium and damage this collagen network and prevent normal primary septation and arrest or distort future lung development. With severe CLD, distal air space diameter increases. There is a failure of primary and secondary septation, arrested lung development and remodeling, with thickened cnt and remodeling, with thickened collagenous saccular walls, and a wide interstitium with increased quantity and size of collagen fibers that can affect the mechanics of ventilation. We conclude that normal lung development is dependent on a normal interstitium and, perhaps, collagen architecture and that origins of CLD begin early in the course of positive pressure ventilation.