Main

NEB are clusters of innervated PNEC that produce amines and peptides(1, 2). Distributed throughout the airway mucosa, PNEC and NEB might play an important role in lung development(3–7) and during neonatal adaptation(3, 4) as NEB are transducers of the hypoxic stimulus and could, therefore, act as airway chemoreceptors in the regulation of respiration(8).

The principal peptide produced by PNEC is GRP, the mammalian counterpart of bombesin(4). Antibodies against BLP or GRP are most widely used as a marker of PNEC in human lungs, because BLP immunopositive cells have been identified in fetal lungs from 7 to 10 wk of gestation onward(2, 9, 10). BLP immunoreactive PNEC differentiate in a craniocaudal direction, and the highest number of cells is found in the small peripheral airways toward the end of gestation(3, 10). Experimental studies revealed that BLP regulates lung branching morphogenesis(7, 11) and stimulates lung growth and maturation(6). A recent study revealed that in mammalian lung the expression of the GRP receptor is developmentally regulated and that the GRP receptor plays an important role, especially during the canalicular stage of lung development(12).

Infants with CDH have abnormal morphologic development of lungs and intrapulmonary blood vessels(13, 14). The high neonatal mortality and morbidity in these children are ascribed to the extent of lung hypoplasia and persistent pulmonary hypertension(15). We have previously reported the expression of CGRP-positive PNEC in a rat model of CDH(16). Lungs of full-term rat pups with CDH contained increased numbers of CGRP-immunostained PNEC compared with the lungs of controls. In the rat, CGRP immunoreactivity has been studied during normal lung development and has been proven to be a reliable marker of PNEC(17), whereas no BLP immunostaining can be detected in this animal species(18). In the human lung the reverse is true. CGRP immunoreactivity has been reported from gestational wk 20 onward(19), but inconsistently and only in a limited number of cells(18, 19).

In this study, using morphometric methods, we investigated the expression of BLP immunostaining in PNEC of lungs from patients with CDH and compared them to lungs from infants with lung hypoplasia due to other causes, and to lungs from control infants without lung abnormalities who died during the perinatal period.

METHODS

Patients. Cases of CDH and of lung hypoplasia due to other causes were selected from the autopsy files of the Departments of Pathology in a large Pediatric Center in Canada (The Hospital for Sick Children in Toronto), and in 10 different hospitals in the Netherlands, spanning the period from 1967 to 1995. Age-matched controls without lung abnormalities were selected from the files of The Hospital for Sick Children in Toronto. The cases for this study were selected on the basis of the clinical diagnosis, of the fixative that had been used (only formalin-fixed tissues were examined), and of good histologic preservation of lung tissue with the presence of intact airway epithelium to identify PNEC and NEB. The use of artificial ventilation for a longer period with high inspiratory peak pressures, especially in cases of CDH, leads to diffuse epithelial damage and hence precludes examination of this lung cell population. Consequently, cases with severe epithelial damage were excluded. Lung hypoplasia was established according to the lung body weight ratio, using the criteria of Askenazi and Perlman(20). Most CDH patients were born at term, and cases from both other groups were therefore selected to obtain the best possible match for gestational age. Control cases were further selected to obtain the best possible match for age at death. Thus, 10 CDH patients were included, as well as seven children with lung hypoplasia and four controls (see Tables 1–3). The autopsy records of all patients were examined for the presence of brainstem pathology.

Table 1 Congenital diaphragmatic hernia: clinical data and results
Full size table
Table 3 Control cases: clinical data and results
Full size table

Histologic examination. Routine 4-μm sections of formalin-fixed, paraffin-embedded lung tissue were immunostained for BLP using the indirect avidin biotin complex staining procedure as previously described(21). All sections were digested with 0.5% pepsin (Sigma Chemical Co.) and incubated overnight with the primary MAb (dilution 1:800) against BLP (Boehringer Mannheim, Germany). Counterstaining was performed with hematoxylin.

With the aid of a projecting microscope (magnification ×700), the total area of airway epithelium of 20 noncartilaginous airways per section and the BLP-immunostained epithelium of these airways were traced on paper(22). The same procedure was done for the 20 largest NEB, located mainly in the medium sized airways, in each section. All drawings were scanned at similar brightness and contrast level, using a Hewlett Packard Scanjet connected to an Apple Macintosh computer. Morphometric analysis included measurements of the total epithelial surface area of 20 airways containing BLP-immunostained cells (referred to as immunostained airways), the BLP-positive areas of airway epithelium in these 20 airways, and the surface area (size) of 20 bronchial NEBs, using the Apple Macintosh National Institutes of Health Image 1.53 program. The BLP-immunostained area in relation to the total epithelial area, referred to as the%IMS-epithelium(22) was calculated from the resulting data. In addition, the%IMS-airways was determined for all sections by counting(23). The average%IMS-epithelium, NEB size, and%IMS-airways were determined per section.

All available sections per case (range 2-5, 62 in total) were studied; 15 sections contained less than 20 noncartilaginous airways to determined the%IMS-epithelium. The median number of airways studied in these sections was 15 (range 6-19). The%IMS-epithelium per case was measured in a median of 40 airways in the CDH group (range 35-80), of 60 airways for the lung hypoplasia group (range 40-85), and of 60 airways (range 52-60) for the controls. Twenty bronchial NEBs could be obtained in all sections. The mean values from the different sections of each case were calculated and compared with other cases.

Data analysis. All data presented are median (range). Differences between groups were tested by one-way ANOVA with the Student-Newman-Keuls test for multiple comparisons or by the nonparametric Kruskal-Wallis test if appropriate. The relationship between clinical data and morphometric results was studied by least square regression. For statistical analysis of the morphometric data, two prematurely born infants (one with CDH and one with lung hypoplasia without CDH; cases 10 and 17, respectively), and two other patients with CDH (one with multiple congenital anomalies, and one with prolonged artificial ventilation; cases 8 and 9, respectively) were excluded to obtain homogeneous groups. Statistical significance was assumed at a two-tailed 5% level.

RESULTS

The clinical data are shown in Tables 1–3. Gestational age, birth weight, and age at death were not significantly different between the three groups; the lung-body weight ratio was significantly higher in control subjects than in both other groups (p < 0.001), whereas the lung hypoplasia cases without CDH had a higher lung-body weight ratio than did CDH patients (p < 0.05).

BLP immunostaining was positive in all sections studied. Qualitative analysis of immunostaining revealed a variable intensity of positive immunostaining. The presence of intensely immunostained NEB was observed in six CDH cases (cases 2, 3, and 4-7), in two cases of lung hypoplasia (cases 11 and 17), and in one control case (case 19). Brown, moderately immunostained NEB were found in one CDH case (case 8), two cases of lung hypoplasia (case 15 and 16), and in three control cases (case 18, 20, and 21). Pale-stained NEB were found in three CDH cases (case 1, 9, and 10) and in three lung hypoplasia cases (case 12-14). In three cases of CDH (case 2, 3, and 7) large NEB, sometimes located “beneath” the epithelium, were found in the large airways and at the bronchoal-veolar junctions (Figs. 1 and 2). This phenomenon was not observed in lung hypoplasia cases (Fig. 3) or in control subjects (Fig. 4).

Figure 1
figure 1

Contralateral lung from a CDH patient (case 3) with severe lung hypoplasia and large NEB located at bronchoalveolar junction (arrow). Bar = 50 μm.

Full size image
Figure 2
figure 2

Ipsilateral lung from a CDH patient (case 2) showing some large NEBs that seem to be located “beneath” the airway epithelium (arrows). Bar = 50 μm.

Full size image
Figure 3
figure 3

Lung hypoplasia case (case 16) showing NEBs within the epithelium of a peripheral airway and bronchoalveolar junction (arrows). Bar = 100 μm.

Full size image
Figure 4
figure 4

Lung from control patient (case 20) showing PNEC and NEBs within the epithelium of a peripheral airway (arrow). Scalebar represents 100 μm.

Full size image

Morphometric data in CDH cases were similar for the ipsilateral and contralateral lungs. Therefore, data from all lung sections of each case were averaged. Statistical analysis of 17 cases (seven CDH cases, six lung hypoplasia cases, and four control cases) of both lungs together showed a higher%IMS-airways in controls than in CDH patients (95 (86-97)% versus 80 (67-85)%, respectively; p = 0.02). The NEB size was significantly larger in lungs of infants with CDH compared with the other two groups (467 (293-656) μm2 in CDH versus 327 (252-472) μm2 in lung hypoplasia, and 312 (215-339) μm2 in controls; p = 0.02). The lung-body weight ratio correlated positively with the%IMS-airways (p = 0.05) and negatively with the NEB size (p = 0.02). The%IMS-epithelium was not significantly different between the groups. However, some cases of CDH with large NEB (Table 1, cases 2, 3, 5, and 7) had also a high value for%IMS-epithelium (Fig. 5).

Figure 5
figure 5

The mean NEB size and the%IMS-epithelium of the ipsilateral and contralateral lungs are shown for CDH patients included in statistical analysis (n = 7); the three CDH patients who were excluded from statistical analysis include cases with multiple congenital anomalies (case 8), prolonged exposure to hyperoxia (case 9), and prematurity, prolonged artificial ventilation, and hyaline membrane disease (case 10); lung hypoplasia cases (n = 6); one prematurely born infant was excluded (case 17), and controls (n = 4). For CDH patients the data from the ipsilateral and contralateral lungs are shown separately; from some slides it was not clear whether they represented the ipsilateral or the contralateral lung. These are indicated as separate symbols (CDH, side unknown). For both other groups the mean value of both lungs is shown. The dashed lines indicate the median NEB size and median%IMS-epithelium of the control cases.

Full size image

Statistical analysis of the separate left and right lungs could be performed in 14 cases-five CDH cases, [three with left-sided CDH (cases 1, 2, and 4), one with right-sided CDH (case 3), and one with bilateral CDH (case 5)], five lung hypoplasia cases (cases 12-16), and four controls. The results are shown in Table 4. No significant differences were observed between the groups.

Table 4 Bombesin-like peptide immunostaining in left and right lungs
Full size table

DISCUSSION

We found that the expression of BLP immunostaining in GRP in lungs of patients with CDH was different from that in lungs of infants with lung hypoplasia due to another cause, and also differed from those without pulmonary abnormalities. Controls had a higher%IMS-airways, which was also reflected in the positive correlation between the%IMS-airways and the lung-body weight ratio. In some CDH cases very large NEBs were found concomitant with a high percentage of immunostained epithelium.

The BLP immunostaining in PNEC and NEB of lungs of infants and children has been investigated for several different pulmonary diseases(2). Only few data on BLP immunostaining in lungs of patients with hypoplastic lungs are available. Absent or very low BLP immunostaining has been reported in preliminary studies of a small number of infants with lung hypoplasia or CDH(4, 24). Jaramillo et al.(25) reported that the density of GRP-immunoreactive cells in lungs of children with anencephaly and lung hypoplasia was similar to that of anencephalic patients without lung hypoplasia and that of normal control subjects. But more GRP-positive cells were located in the airways in anencephalic patients with lung hypoplasia than in those of patients without lung hypoplasia(25). It is not clear whether differences in tissue processing, methodology, or the gestational ages might explain the differences with our observations.

An important prerequisite for our study was the preservation of airway epithelium. Therefore, patients with extensive epithelial damage who had been ventilated for more than a few hours had to be excluded. This resulted in a selection of CDH patients with severe lung hypoplasia resulting in death shortly after birth. Most children had died within the first 2 h after birth, and only two children had lived for more than 2.5 h. The number of cases studied was the maximal number that could be retrieved from files of the Pathology Departments of 10 institutions in the Netherlands and in The Hospital for Sick Children in Toronto between 1967 and 1996. (Before this period, Zenker's fixative has been used, precluding immunostaining for BLP). This represents the first study of BLP immunostaining in a comprehensive cohort of infants with CDH.

The variation in BLP immunostaining observed in CDH cases might be partly explained by the differences in clinical history. The lungs of the patients who survived longest (cases 9 and 10) and of a dysmaturely stillborn patient with multiple congenital anomalies (case 8) showed pale immunostaining, relatively small NEB, and the lowest%IMS-epithelium. Patient number 10 was prematurely born and had hyaline membrane disease, a condition that is known to decrease BLP immunostaining(4, 9, 26). It can be assumed that the prolonged treatment with artificial ventilation and oxygen therapy in cases 9 and 10 and the renal abnormalities in case 8 might have influenced the BLP immunostaining in the lungs. These cases were not included in the statistical analysis, to maintain uniform clinical parameters.

The largest NEB and the highest%IMS-epithelium were found in lungs from CDH patients with severe hypoplasia who could not be ventilated adequately and died shortly after birth. Whether the function of these large NEB is abnormal is speculative. It is known that NEB are transducers of the hypoxic stimulus(8), and that increased exocytosis with secretion of neurotransmitters occurs during hypoxia(27). We speculate that large NEB might indicate failure of neuropeptide release, i.e. that the hypoxic secretory response has not occurred. Such a phenomenon has been reported in infants with asphyxia and loss of brainstem function who generally died after more than 24 h(28). In the present study the brainstem was examined in most infants, and no evidence of brainstem damage was observed. This could, therefore, not explain our findings.

On the other hand, it can be assumed that, during resuscitation of the CDH patients using artificial ventilation and a high inspired oxygen fraction, the local concentration of oxygen in the airways was high, which might have inhibited exocytosis. This is supported by the findings of Lauweryns et al.(29), who reported that a low oxygen concentration of the inhaled air, but not hypoxemia, stimulates secretion of serotonin by NEB.

The lungs of children with lung hypoplasia, secondary to renal or genitourinary abnormalities, showed variation in BLP immunostaining. The pattern of BLP immunostaining was, however, different from that in CDH patients. Half of the cases with lung hypoplasia showed pale immunostaining, and the NEB size and the%IMS-epithelium were within the same range as that found in control subjects. The differences in BLP immunostaining between the lung hypoplasia cases could not be explained by differences in gestational age or in clinical presentation.

In conclusion, we found that the BLP immunostaining in lungs of CDH patients differs from that of age-matched controls and from that of infants with lung hypoplasia due to other causes. Infants with the most severe lung hypoplasia and persistent pulmonary hypertension, as reflected by their very early age at death, have increased BLP immunostaining. Because BLP stimulates lung growth and lung maturation(6) it might be assumed that the increased BLP immunostaining reflects a maximal response of the lung to compensate for the abnormal growth in CDH. In an experimental setting using a rat model of CDH enlarged NEB and increased expression of CGRP have been reported(16, 30). In the developing rat lung, CGRP immunoreactivity reaches its maximum near term(17), a developmental pattern similar to that of BLP in the human lung(3). We therefore propose that the rat model of CDH is suitable for the performance of further studies of the altered expression of neuroendocrine cells and their peptides in the lungs in CDH.

It can also be assumed that the increased BLP immunostaining, observed in some CDH patients, reflects the extent of persistent pulmonary hypertension. This assumption is supported by a study that reported increased BLP immunostaining in older patients with primary pulmonary hypertension(31), although BLP itself exerts no effect on the pulmonary vascular tone(32, 33). The large BLP-positive NEBs in lungs of CDH patients might, however, contain other peptides, such as leuenkephalin(21), endothelin(2), or serotonin(2), which are known to induce pulmonary vasoconstriction(2, 32). However, our present data do not allow for conclusions regarding the role of NEB in abnormal lung development in CDH.

Table 2 Lung hypoplasia: clinical data and results
Full size table