American Journal of Respiratory and Critical Care Medicine

We present the clinical, radiologic, and pathologic findings in lung biopsies from seven infants with atypical neonatal lung disease. All seven infants presented with tachypnea, hypoxemia, and diffuse interstitial infiltrates with overinflated lungs on chest radiographs in the first month of life. Lung biopsies from all cases showed similar pathology, with expansion of the interstitium by spindle-shaped cells containing periodic acid–Schiff positive diastase labile material consistent with glycogen. Immunohistochemical staining showed these cells to be vimentin positive but negative for leucocyte common antigen, lysozyme, and other macrophage markers. Electron microscopy revealed primitive interstitial mesenchymal cells with few cytoplasmic organelles and abundant monoparticulate glycogen. Minimal or no glycogen was seen in the alveolar lining cells. Five cases were treated with pulse corticosteroids; hydroxychloroquine was added in one case. Six of seven infants have shown a favorable clinical outcome. One infant died from complications of extreme prematurity and bronchopulmonary dysplasia. Three cases that have been followed for at least 6 years have shown clinical resolution and radiographic improvement. We propose the term “pulmonary interstitial glycogenosis” of the neonate for this new entity to be differentiated from other forms of interstitial lung disease. Because abundant glycogen is not normally found in pulmonary interstitial cells, we postulate an abnormality in lung cytodifferentiation involving interstitial mesenchymal cells.

Interstitial lung disease (ILD) is an uncommon cause of respiratory distress during the perinatal period. The various types of interstitial pneumonitis occurring in infancy and childhood have been reviewed (13). The classification schemes used in these studies are based mostly on adult ILD; a partial list includes usual interstitial pneumonitis (UIP), desquamative interstitial pneumonitis (DIP), lymphoid interstitial pneumonitis (LIP), and nonspecific interstitial pneumonitis. Fan and Langston (1) divided chronic ILD in the pediatric age group into two broad categories. The first category included entities with known etiology (i.e., infections, oxygen toxicity, environmental agents), whereas the second category consisted of cases of chronic ILD of unknown etiology. Although the general histologic patterns of interstitial pneumonitis in adults and children share many similarities, specific subtypes seen only in the pediatric age group, including “chronic interstitial pneumonitis of infancy” (CPI) (4) and “cellular interstitial pneumonitis of infants” (CIP) (5), have been described. The etiology of both CPI and CIP is unknown. Whereas CPI is believed to reflect a slowly resolving or recurrent pneumonia superimposed on immature or abnormally developed lung (4), CIP was postulated to represent an autoimmune process with persistent histiocytic inflammation triggered by an ephemeral microbiologic insult (5).

Here we report seven cases of chronic ILD presenting during the neonatal period, in whom all known infectious or inflammatory causes of neonatal lung disease were excluded and lung biopsies showed uniform histologic features characterized by thickened interstitium due to the presence of immature interstitial cells containing abundant cytoplasmic glycogen. We present detailed clinical and laboratory findings, imaging studies, treatment, long-term follow-up, and histopathologic, immunohistochemical and ultrastructural findings in lung biopsies from these patients. On the basis of our observations we propose the term “pulmonary interstitial glycogenosis” because this feature suggests a developmental disorder, in contrast to other types of ILD due to infection or inflammatory etiology. We propose that the underlying defect may involve an abnormality in differentiation of pulmonary mesenchyme because abundant glycogen is not normally found in pulmonary interstitial cells.

Open lung biopsies were performed between 2 weeks and 4 months of age. Samples of lung biopsy tissue were fixed in 10% neutral buffered formalin and embedded in paraffin. For histopathologic assessment, paraffin sections were stained with hematoxylin and eosin (H.E.), periodic acid–Schiff (PAS) with and without diastase treatment, Masson trichrome, and Van Gieson stains for connective tissue elements. Stains for microorganisms included Gram stain, Steiner stain, Grocott methanamine silver, and Ziehl-Neelsen stain for mycobacteria.

For immunohistochemical studies, the indirect immunoperoxidase method was employed on formalin-fixed paraffin sections. The following primary monoclonal (MAb) or polyclonal antibodies were used on all biopsies: MAb against low–molecular weight cytokeratin (Mol. nos. 8, 18, 19; Beckton-Dickinson, Mountain View, CA), MAb against actin, the leukocyte marker CD45RB, the macrophage marker CD68, vimentin (all from Dako Corporation, Carpinteria, CA), and polyclonal antibody against lysozyme (Dako).

For transmission electron microscopy (TEM) studies, samples of lung were obtained for all the cases, fixed in “universal fixative” (1% glutaraldehyde and 4% paraformaldehyde in 0.1 M phosphate buffer) post-fixed in 2% osmium tetroxide, and embedded in Epon. One-micron Epon sections were stained with toluidine blue or PAS for demonstrating glycogen. For conventional TEM, ultrathin sections were contrasted with uranyl acetate followed by lead citrate. To demonstrate glycogen at TEM level we used the tannic acid method (6) or the periodic acid–silver methanamine (PA-SM) method as previously reported (7).

To rule out the possibility of glycogen storage disease, skin fibroblasts from Case 2 were analyzed (testing performed by Dr. Y. T. Chen, Genetic Testing Laboratory, Duke University Medical Centre, Durham, NC); they showed normal glycogen content, acid α-glucosidase, and branching enzyme activities.

Case Histories

Two of the patients were lost to follow-up, and one died as a result of significant prematurity and associated bronchopulmonary dysplasia (BPD). The three cases that we have followed for a minimum of 6 years are described in more detail (Table 1)

TABLE 1. Summary of seven cases of pulmonary interstitial glycogenosis


Case

Sex

Birth
 Weight
 (kg)

GA
 (wk)

Pre-/Perinatal Complications

Age of
 Presentation

Mechanical
 Ventilation

Age at
 Lung
 Biopsy

Steroid Use

Age at Last
 Follow-up

Outcome
1 M2.5533Incompetent cervix,
   Shirodhkar suture5 dYes
   (Days 7–12)25 dYes6.5 yrWell; mild intermittent
   bronchospasm
2 M2.9638Bleeding first trimester,
   preeclampsia third trimester
   (Rx: phenobarbital),
   cephalopelvic disproportion,
   bicornuate uterus1 dYes
   (Days 1–4)8 wkYes6 yrWell; mild intermittent
   bronchospasm
3 M3.3040None4 wkNo3 moYes and
 hydroxychloroquine6.5 yrWell
4 M2.0033Cesarian for fetal distress1 dYes
   (Days 1–7)7 wkNo10.5 moOn oxygen. Lost to
   follow-up due to move
5 F1.2429Flulike illness first trimester,
   rubella nonimmune1 dYes
   (Days 2–15)11 dYes26 moWell. Lost to follow-up due
   to move
6 M5.0840Breech,
   cephalopelvic disproportion
   cesarian1 dNo19 dNo11 yrWell. (Lost to follow-up,
   last seen in clinic at 7 mo,
   but seen in ER at 11 yr)
7
 M
0.82
25
Maternal chronic sinusitis,
   placental abruption
1 d
Yes
   (Days 0–49,  84–126, then
   intermittently
   until 6 mo)
24 d
Yes
6 mo
Died. Cor pulmonale

Definition of abbreviations: ER = emergency room; F = female; GA = gestational age; M = male.

.

Case 1.

This male infant was born to a 34-year-old gravida 4 (total number of pregnancies), para 1 (number of pregnancies reaching viability), and aborta 2 (number of abortions) mother at 33 weeks gestational age following a pregnancy complicated by an incompetent cervix, requiring placement of a Shirodhkar suture at 26 weeks gestation. The suture was removed at 33 weeks gestation, and spontaneous onset of labor ensued. The infant was born by spontaneous vaginal delivery and had a birth weight of 2,550 g, with APGAR scores of 7 and 9 at 1 and 5 minutes, respectively. Respiratory distress developed on Day 5 of life; supplemental oxygen was initiated, and the infant was transferred to a tertiary care hospital. He experienced gradual deterioration in his respiratory status and subsequently required assisted ventilation from Day 7 of life for a total of 5 days. This was followed by 16 days of nasal continuous positive airway pressure. Initial chest radiographs demonstrated large volume lungs with a fine interstitial pattern, worse in the right lung (Figure 1a)

.

Various investigations, including echocardiograms and a pulmonary angiogram, were normal. Blood, urine, and deep endotracheal tube cultures were negative. There was no clinical improvement after a 7-day course of ampicillin and amikacin and a 14-day course of erythromycin. At the age of 25 days the infant underwent open lung biopsy, which revealed diffuse interstitial hypercellularity. Neither histologic examination nor culture of the lung tissue demonstrated infectious organisms.

The infant had an uneventful course postbiopsy and was discharged 9 days later from the referring hospital, at which time his capillary blood gas tensions and acid base status were as follows: carbon dioxide tension (Pco2) 46 mm Hg; pH 7.39; HCO3 27 mmol/L, with a base excess of 1.8 mmol/L. When seen the next day at our institution for the first time, he was not in acute distress and his height and weight were above the 50th percentiles when corrected for his gestational age. He did not have adventitial sounds on auscultation, but intercostal retractions were present. Pulse oximetry revealed an oxyhemoglobin saturation of 90–92% while breathing room air. Intravenous pulse corticosteroid therapy (methylprednisolone, 10 mg/kg/day once a day for 3 days each month) was initiated, and within 3 months his oxyhemoglobin saturation on room air had improved to 97%. Pulse corticosteroid therapy was discontinued after a total of six monthly treatments. The subject has subsequently remained well except for a brief hospitalization for a viral upper respiratory tract infection and bronchospasm at 2.5 years of age that quickly resolved with bronchodilators, corticosteroids, and oxygen. A chest radiograph done at the age of 5.5 years (Figure 1b) revealed large lung volumes. He had an inspiratory and expiratory high-resolution computerized tomography (HRCT) scan of the chest (Figures 1c and 1d), which showed normal lung parenchyma and absence of air trapping. There was some mild lingular atelectasis and scarring in the right middle lobe, believed to be secondary to the site of lung biopsy. He is now 6.5 years old, active, asymptomatic, and has a normal physical examination, and his oxyhemoglobin saturation is 99% on room air. Presently he requires intermittent bronchodilator and inhaled corticosteroid therapy during viral respiratory tract infections.

Case 2.

This 2.96 kg male infant was born to a 32-year-old gravida 1, para 0 mother at 38 weeks gestation by cesarean section after spontaneous onset of labor. The pregnancy was complicated by bleeding in the first trimester and preeclampsia in the final month, during which period she received phenobarbital. Although cephalopelvic disproportion and a bicornuate uterus complicated the delivery, the APGAR scores were 8 and 9 at 1 and 5 minutes, respectively. At 3 hours of age, he was noted to be tachypneic and to have intercostal indrawing, and supplemental oxygen therapy was begun (FiO2 0.28–0.35). Arteriolized capillary blood gas tensions and acid base status revealed Pco2 to be 50 mm Hg, oxygen tension (Po2) 42 mm Hg, pH 7.29, HCO3 24 mmol/L and a base excess of −2 mmol/L. Initial chest radiographs demonstrated a hazy bilateral mixed interstitial and alveolar pattern with normal volume lungs (Figure 2a)

. A left-sided chest tube was placed to decompress an initial left pneumothorax. The subject continued to deteriorate, ultimately requiring intubation and transfer to our institution, where he received assisted ventilation for 4 days and continuous positive airway pressure for 2 days. Before transfer, antibiotics were begun, and he subsequently received a 7-day course of ampicillin and gentamicin and a 9-day course of erythromycin. Although maternal vaginal and cervical swabs were positive for Ureaplasma urealyticum, the cultures of blood and tracheal aspirates, including ureaplasma, were negative. Serology for cytomegalovirus, toxoplasmosis, mycoplasma, herpes simplex virus, rubella, and human immunodeficiency virus were negative. An echocardiogram revealed a small patent foramen ovale. A barium swallow showed significant gastroesophageal reflux, with no evidence of aspiration.

At 7 days of age the subject was transferred back to the referring hospital with minimal intercostal retractions, a respiratory rate of 60–70 breaths/minute, and a requirement for supplemental oxygen therapy (FiO2 = 0.3). His oxygen dependency, respiratory distress, and abnormal chest radiographic findings persisted despite a second 7-day course of ampicillin and gentamicin, and a 21-day course of erythromycin. During this time he had a transient eosinophilia (18%). His height and weight were at the 10th and 25th percentiles, respectively, at 8 weeks of age. He was transferred back to our institution, and at 8 weeks of age he underwent an open lung biopsy. The routine histology on lung biopsy showed features of CIP with diffuse interstitial thickening and hypercellularity (Figures 3a and 3b)

, with no evidence of infection on culture and histologic examination of the lung.

Intravenous pulse corticosteroid therapy (methylprednisolone, 10 mg/kg/day once a day for 3 days each month) was then initiated. The subject required supplemental oxygen therapy for 3 months after biopsy, at the end of which time his oxyhemoglobin saturation on room air had improved to 97% and has since remained normal. He received a total of 6 months of pulse corticosteroid therapy, after which his height and weight were at the 75th percentile. Chest radiographs at 6 months age (Figure 2b) demonstrate significant clearing of the previous interstitial densities; however, the lung volumes remain overinflated, with some residual coarse basal interstitial changes. A computerized tomography scan performed at 7 months age (Figure 4)

demonstrated some coarse linear bands of opacity, likely representing scars, and some dependent nonspecific ground glass density.

The subject has continued to do well and his only subsequent therapy has been intermittent infrequent inhaled corticosteroids and bronchodilators for wheeze during viral respiratory tract infections. At 6 years age, he is asymptomatic, with good exercise tolerance, and his physical examination is normal. His oxyhemoglobin saturation on room air is 98–100%, forced vital capacity 90%, forced expiratory volume in 1 second 78%, and mean forced expiratory flow during the middle half of the forced vital capacity 48% of predicted values. These values improved significantly after administration of a bronchodilator to 96%, 87%, and 60%, respectively, suggestive of reversible obstructive airways disease. The chest radiograph remains unchanged, demonstrating mildly overinflated lungs.

Case 3.

This male infant was born after a normal pregnancy by spontaneous vaginal delivery at 40 weeks gestation, with a birth weight of 3,280 g. The postpartum period was uneventful, and he was discharged home at 2 days age. At 1 month of age, the parents were concerned that he was frequently “congested,” with episodes of “noisy breathing.” At 7 weeks of age the infant was admitted to our hospital with fever and subcostal retractions, and a chest radiograph revealed a diffuse and fairly severe mixed interstitial and alveolar pattern, worse at both bases (Figure 5a)

. The lung volumes were increased. Intravenous ampicillin and gentamicin were started. Blood, urine, and cerebrospinal fluid cultures were negative. The infant was discharged home on oral amoxicillin with an oxyhemoglobin saturation of 95% in room air.

At 10 weeks of age he presented again with a several-weeks history of tachypnea and a 2-day history of fever, lethargy, and poor appetite. His sleeping respiratory rate was 40 breaths/minute, there was intercostal indrawing, and bilateral crackles were heard on auscultation of lung fields. The infant's height was at the 75th percentile, and his weight was at the 10th percentile. While inspiring room air his oxyhemoglobin saturation was 70%, and capillary blood analysis revealed the pH to be 7.44, Pco2 37 mm Hg, and HCO3 24 mmol/L. His complete blood cell count revealed a white cell count of 15.5 × 109cells/L, with a slight predominance of neutrophils. Analyses of his cerebrospinal fluid, sweat chloride concentration, thyroid function, and immunoglobulin concentration were normal. Blood and urine cultures, nasopharyngeal immunofluorescence studies for viral infection, and serology for human immunodeficiency virus were all negative. Other studies, including an electrocardiogram, barium swallow, and feeding study, were unremarkable.

At the age of 3 months he underwent an open lung biopsy because of persistence of respiratory symptoms and requirement for supplemental oxygen. The findings on lung biopsy were those of mild to moderate interstitial thickening and hypercellularity. As in the other cases, no infective agents could be identified on culture and histologic examination of the lung.

Oral hydroxychloroquine was initiated at the age of 5 months and discontinued at 7 months, when no improvement was seen. At 6 months of age he was started on monthly courses of high-dose pulse corticosteroids, as described previously, which he received until the age of 16 months. Subsequent chest radiographs at ages 12 months and 2 years (Figure 5b) demonstrate significantly overinflated lungs and persistent bi-basal interstitial changes. The interstitial disease, although persistent, was significantly improved over earlier films. A HRCT scan at 2 years of age (Figure 6)

demonstrated large areas of patchy basal and central dense ground glass opacities.

His course was complicated with episodes of presumed fluid overload requiring oral diuretics (spironolactone and hydrochlorothiazide), frequent brief hospitalizations for viral upper respiratory tract infections, and the requirement for gastrojejunal feeds via gastrostomy. At 9 months of age he had a right middle lobe pneumonia requiring admission to hospital for intravenous antibiotics. Diuretics were discontinued at 15 months of age, and at 16 months he had an oxyhemoglobin saturation on room air of 93% and supplemental oxygen was discontinued. He continued to have frequent viral upper respiratory tract infections, with intermittent episodes of bronchospasm and requiring two brief admissions at the age of 2.5 years. He was treated with inhaled corticosteroids, bronchodilators, and oxygen. At 3 years and 9 months of age, he developed a left lower lobe pneumonia from which he recovered. By 6 years of age he was no longer symptomatic and had a normal exercise tolerance. Although auscultation of lung fields revealed diffuse crackles bilaterally, his respiratory rate was 16 breaths/minute and his oxyhemoglobin saturation was 98% while breathing room air. Chest radiographs at 10 years of age (Figure 5c) demonstrate a pattern similar to the previous cases, with significant bilateral overinflation and coarse basal interstitial changes, worse in the region of the posterobasal segments of the lower lobes.

Light Microscopy

Conventional histologic examination of lung biopsies from all the cases showed similar changes. The most striking abnormality was diffuse interstitial thickening without apparent hyperplasia of alveolar lining cells or marked accumulation of cells or proteinaceous material within the alveolar spaces. (Figure 3a). On higher magnification the interalveolar septae were seen to be uniformly expanded by a population of oval- to spindle-shaped cells with pale cytoplasm, indistinct cell membrane, and oval, bland nuclei (Figure 3b). Routine sections stained with PAS showed patchy PAS positive diastase sensitive material within the cytoplasm of these interstitial cells, indicating the presence of glycogen. However the preservation of glycogen varied between cases because of the use of aqueous fixative (i.e., 10% formalin). The preservation of glycogen in interstitial cells was markedly improved in lung tissue fixed for electron microscopy (EM) studies. High-resolution microscopy on a 1-μm Epon section stained with PAS showed uniform positive cytoplasmic staining (Figure 3c). The interalveolar septae contained occasional inflammatory type cells but there was no evidence of interstitial fibrosis. The airways and blood vessels represented in the biopsies showed no obvious abnormalities. All special stains for microorganisms were negative.

Immunostaining for cytokeratin outlined the alveolar lining cells distributed singly or in small clusters (Figure 7a)

. The population of interstitial cells showed positive immunoreactivity for vimentin (Figure 7b) and focal positivity for actin (not shown), whereas other immunostains including CD45 (common leukocyte antigen) (Figure 7c), CD68, and lysozyme (macrophage markers) (Figure 7d) were all negative. In the same sections, occasional inflammatory type cells and/or alveolar macrophages showed positive reactions with the aforementioned markers (positive internal control).

EM

At ultrastructural level, the population of interstitial cells showed features of primitive mesenchymal cells with a paucity of cytoplasmic organelles. In routine TEM samples, the cytoplasm of these cells appeared empty or contained low contrast granular material suggestive of glycogen (Figures 8 and 9)

. In ultrathin sections treated with tannic acid method or PA-SM, well-defined electron-dense particles typical of nonparticulate glycogen (or so-called β particle glycogen) were clearly visualized (Figures 10 and 11) . Some interstitial cells in addition to glycogen also contained occasional droplets of neutral fat, but there were no significant aggregates of intermediate filaments, cytoplasmic granules, or lysozomelike bodies (Figure 9). The degree of cytoplasmic differentiation of interstitial cells varied (Figure 10). Some interstitial cells, although still containing dispersed cytoplasmic glycogen, also showed more prominent organelles, particularly rough endoplasmic reticulum forming anastomosing cinternae, characteristic of fibroblasts (Figure 11).

The other cell types normally found within lung interstitium, i.e., vascular constituents (endothelial cells, pericytes), histiocytes, and leukocytes, exhibited the usual ultrastructural features without cytoplasmic glycogen. The alveolar lining cells consisted of Type I cells with thin cytoplasmic processes forming air–blood barriers and cuboidal Type II cells with surface microvilli and characteristic cytoplasmic lamellar bodies corresponding to surfactant storage sites. In alveolar Type II cells, occasional patches of cytoplasmic glycogen were present, but their overall ultrastructural appearance was that of well-differentiated alveolar epithelial cells. No significant pools of cytoplasmic glycogen were observed in the other lung cell types represented in the lung biopsy samples including airway epithelium, airway and vascular smooth muscle, or endothelial cells.

Because of this unusual ultrastructural observation of primitive glycogen-rich mesenchymal-like interstitial cells in our seven cases, we reviewed other cases of neonatal lung disease, normal age-matched control subjects, and fetal lungs at different stages of development. No cases with the aforementioned features were found among over 1,000 cases of pediatric lung biopsies with TEM studies. These cases included various types of ILD (i.e., DIP, UIP, LIP) cases of surfactant protein B deficiency, congenital deficiency of lamellar bodies (8), alveolar capillary dysplasia, and acinar dysplasia as well as cases of pulmonary prematurity with and without hyaline membrane disease, BPD, and the syndrome of Wilson and Mikity (9). Primitive interstitial cells with abundant cytoplasmic glycogen were not seen in any of the conditions noted in the foregoing. In normal human fetal lungs from early (6 weeks) to late (30 weeks) gestation, the epithelial cells contained glycogen as documented well earlier (10), but no glycogen was observed in interstitial mesenchymal cells at any gestational age.

Our case series outlines the clinical findings and long-term follow-up of children who initially presented with atypical noninfective respiratory disorders during the neonatal period. Although there was some variation in the rate of clinical improvement, six of the seven infants showed an excellent clinical outcome. This contrasts with the overall mortality rate of 21% in pediatric ILD (11) and UIP mortality rates of 7–50% (12, 13).

LIP, DIP, and UIP are characterized by an inflammatory process (1, 14). Although UIP is the most frequent form of ILD seen in adults, it rarely occurs in children. DIP, believed by some to be an earlier stage of UIP, is more common during infancy. Although LIP is most commonly associated with acquired human immunodeficiency syndrome, it can also be an idiopathic disorder (15). Schroeder and colleagues (5) first differentiated CIP from classic ILD. Their five cases presented with tachypnea in the first 2 weeks of life and had bilateral interstitial infiltrates, and all but one patient recovered 4–18 months after diagnosis, when they were 2–2.5 years of age. Eight additional cases of CIP have been reported (16). Familial ILD associated with a mutation of the surfactant protein C gene has been described in a 1-year-old girl and her mother (17). The conventional radiographs of these children in our series, although nonspecific, are remarkably similar. Unlike what is seen in early respiratory distress syndrome, these children seem to present with a diffuse fine interstitial reticular pattern in the setting of overinflated lungs. In some films the interstitial pattern approaches the type of ground glass density seen in hyaline membrane disease, but this pattern on a background of overinflated lungs resembles early diffuse infectious pneumonia. However, the pattern seems to progress rapidly to a coarser interstitial pattern characterized by areas of linear coarse opacities mixed with overinflated/emphysematous zones, worse at the bases. This pattern resembles the sequence of radiographic changes that occur in traditional Northway–Edwards Stage IV BPD (18). The descriptions of the radiographic manifestations of this disorder to date have been minimal (5, 16), but those that are available are congruent with our present radiographic observations.

The most striking finding in the lung biopsies from our seven patients was diffuse, uniform interstitial thickening due to the presence of immature interstitial cells containing abundant cytoplasmic glycogen. To the best of our knowledge this lung pathology has not been described previously and may represent a unique form of neonatal ILD. On the basis of the lung biopsy histology, negative special stains for microorganisms, and negative culture studies, the possibility of an infectious etiology has been excluded in all our cases. There was only a minor component of interstitial inflammatory cells, as revealed by immunohistochemistry. This may represent a basal contingent of lymphoid cells rather than a pathologic process. The pulmonary vasculature and peripheral airways revealed no abnormalities. Because there was no abnormal accumulation of macrophages and marked alveolar Type 2 cell hyperplasia was absent, the patient having DIP is an unlikely diagnosis. Similarly there were no histologic features to suggest UIP or LIP in our cases. The CPI described by Katzenstein and coworkers (4) is characterized by marked alveolar septal thickening, alveolar pneumocyte hyperplasia, and intra-alveolar exudate containing numerous macrophages and some eosinophilic debris. The etiology of CPI is believed to be multifactorial, representing a slowly resolving or recurrent pneumonia superimposed on immature or abnormally developed lung (4). In addition, the clinical course and prognosis of reported cases of CPI appears more ominous compared with our cases. The histologic findings in lung biopsies from the cases of CIP reported by Schroeder and associates (5) bear a striking resemblance to our cases of pulmonary interstitial glycogenesis. Both the original paper (5) and a recent case report of CIP (16) describe an interstitial monocellular infiltrate widening of the alveolar walls associated with mild alveolar Type 2 cell hyperplasia but absent in interstitial fibrosis or significant airway disease. The immunohistochemistry in these cases has been reported to show an excess of lymphocytes positive for leukocyte common antigen and interstitial cells positive for α-antichymotrypsin, suggesting a histiocytic origin. However, special stains for glycogen or EM were not performed in these previous reports (5, 16). Although the etiology of CIP has not been defined, the authors suggested that this lesion may represent an autoimmune process with persistent histiocytic inflammation. It remains to be determined whether CIP may in fact represent cases of pulmonary interstitial glycogenosis.

Although the lung pathology in our cases was suggestive of CIP, the precise nature of the underlying abnormality was not appreciated on routine histology. The finding of immature interstitial cells containing abundant cytoplasmic glycogen became apparent only in lung tissue samples processed for EM studies. Some immature interstitial cells showed aggregates of anastomosing cisternae of endoplasmic reticulum, a feature suggestive of fibroblast differentiation. There were no ultrastructural features to indicate histocytic or inflammatory cell phenotype. These immature interstitial cells were found in lung biopsies from all our cases, irrespective of the birth weight, gestational age (ranging from 25 to 40 weeks), or age at biopsy. This suggests that this lung abnormality originates in utero and that it persists for several months after birth. The immature interstitial cells were mostly located in the thick portion of interalveolar septae and were distributed among other interstitial cell components (i.e., endothelial cells, pericytes, blood cells) that appeared well differentiated. Similarly the alveolar lining cells, including Type 2 cells, appeared well-differentiated, with expected ultrastructural features such as well-developed lamellar bodies, storage sites for surfactant. The cytoplasm of occasional Type 2 cells contained residual aggregates of glycogen particles, as can be seen in premature or even full-term neonatal lungs (10, 19, 20). The presence of glycogen in epithelial cells of developing lung has been well-documented in humans (10) as well as various mammals (19, 20). It has been suggested that glycogen may be an important source of substrate for fatty acid synthesis in fetal lung (21). There appears to be an inverse temporal relationship between glycogen content and biosynthesis of fatty acids and phosphatidyl choline in developing fetal lung (20, 21). At the ultrastructural level, it has been shown that in developing rabbit lung, a decrease in glycogen content in alveolar Type 2 cells coincided with maturation of the surfactant system (19). Therefore it has been suggested that glycogen in alveolar Type 2 cells may provide a substrate for surfactant synthesis (22).

Large cytoplasmic pools of glycogen are not normally found in pulmonary interstitial cells either during fetal development or post-natally. Our study of cytodifferentiation of pulmonary epithelium in developing human lungs (10) as well as ultrastructural studies in primate lungs (23) found no evidence of glycogen accumulation in interstitial cells at any stage of lung development. On the other hand, primitive airway epithelium contained abundant glycogen, ruling out the possibility of glycogen extraction during fixation or processing. In developing rat lungs, glycogen was reported in so-called lipofibroblasts, where small pools of glycogen were noted in association with prominent lipid droplets (24). In the rat, the lipofibroblasts are usually seen during the early post-natal period and are believed to represent precursors for myofibroblasts (25). The morphologic, immunohistochemical, and biochemical characteristics of pulmonary lipofibroblasts have been reviewed recently by McGowan and Torday (26). Pulmonary lipofibroblasts have been most extensively characterized in the rat but have also been described in mouse and hamster lungs (25). However there is only limited information on this cell type in humans (27).

There are neither controlled trials of nor satisfactory guidelines for the treatment of pediatric ILD. The inflammation that characterizes pediatric ILD has led to the use of corticosteroids (1), and a favorable response has been reported in 40% of patients in a retrospective study (13). DIP, relative to UIP, appears to respond more frequently to corticosteroid therapy (1). Chloroquine and hydroxychloroquine (28) have also been used to treat pediatric ILD, and response rates have ranged from 40 to 60% (11, 29). Few adult patients with idiopathic pulmonary fibrosis respond to corticosteroids in the short term, and corticosteroids do not enhance their survival (30).

Primary interstitial glycogenosis, in contrast to UIP, DIP, and LIP, is not characterized by inflammation, and the primary abnormality is dysmaturity of the interstitial cells. Thus, if there is a beneficial effect from corticosteroid therapy, it likely results from an acceleration of the maturation process rather than from modifying inflammation. Some patients with CIP are believed to have benefited from hydroxychloroquine (16) or prednisone in combination with hydroxycholoroquine and azathioprine (5). We did not observe any clinical benefit in the one patient we treated with hydroxychloroquine. When compared with other idiopathic neonatal ILD such as fibrosing alveolitis (12, 29), LIP, UIP, or DIP (1, 13, 14), pulmonary interstitial glycogenosis has a more favorable long-term prognosis. Whether this results from a more favorable response to corticosteroids is unknown.

In conclusion, pulmonary interstitial glycogenosis is a rare neonatal lung disease that is characterized by the presence of glycogen laden cells within the interstitium of the lung and is presumed to be the same entity as CIP (5, 16). There appears to be selective “dysmaturity” of interstitial cells without apparent effects on Type 2 or endothelial cell differentiation or function. The precise etiology and/or pathogenic mechanisms involved in pulmonary interstitial glycogenosis remain unknown. Our findings suggest a developmental abnormality rather than an inflammatory or reactive process. Although the precise molecular mechanisms involved in pulmonary interstitial glycogenosis are unknown, the identification of these cells in lung tissue serves as a useful marker for this disorder. This new terminology will help to differentiate this entity from other ILD in the perinatal period and also emphasizes its apparently more favorable outcome relative to other chronic ILD.

The authors thank L. Lines and V. Edwards for excellent technical assistance with electron microscopy and M. Starr for photography.

Supported in part by grants from Canadian Institute for Health Research, Group on Lung Development (E. C., H. O'B.).

1. Fan LL, Langston C. Chronic interstitial lung disease in children. Pediatr Pulmonol 1993;16:184–196.
2. Nicholson AG, Kim H, Corrin B, Bush A, du Bois RM, Rosenthal M, Sheppard MN. The value of classifying interstitial pneumonitis in childhood according to defined histological patterns. Histopathology 1998;33:203–211.
3. Howenstine MS, Eiger H. Current concepts on interstitial lung disease in children. Curr Opin Pediatr 1999;11:200–204.
4. Katzenstein AL, Gordon LP, Oliphant M, Swender PT. Chronic pneumonitis of infancy: a unique form of interstitial lung disease occuring in early childhood. Am J Surg Pathol 1995;19:439–447.
5. Schroeder SA, Shannon DC, Mark EJ. Cellular interstitial pneumonitis in infants: a clinicopathologic study. Chest 1992;101:1065–1069.
6. Hayat MA. Principles and techniques of electron microscopy: biological applications, 2nd ed. Baltimore, MD: University Park Press; 1981. p. 61–63.
7. Cutz E, Conen PE. Ultrastructure and cytochemistry of Clara cells. Am J Pathol 1971;62:127–134.
8. Cutz E, Wert SE, Nogee LM, Moore AM. Deficiency of lamellar bodies in alveolar type 2 cells associated with fatal respiratory disease in a full-term infant. Am J Respir Crit Care Med 2000;161:608–614.
9. Swyer PR, Delivoria-Papadopoulous M, Levison H, Reilly BJ, Balis JU. The pulmonary syndrome of Wilson and Mikity. Pediatrics 1965;36:374–384.
10. Cutz E. Cytomorphology and differentiation of airway epithelium in developing human lung. In: McDowell EM, editor. Lung carcinomas. Edinburgh: Churchill Livingstone; 1987. p. 1–41.
11. Fan LL. Evaluation and therapy of chronic interstitial pneumonitis in children. Curr Opin Pediatr 1994;6:248–254.
12. Hewitt CJ, Hull D, Keeling JW. Fibrosing alveolitis in infancy and childhood. Arch Dis Child 1977;52:22–37.
13. Fan LL, Mullen ALW, Brugman SM, Inscore SC, Parks DP, White CW. Clinical spectrum of chronic interstitial lung disease in children. J Pediatr 1992;121:867–872.
14. Bokulic RE, Hilman BC. Interstitial lung disease in children. Pediatr Clin North Am 1994;41:543–567.
15. O'Brodovich H, Moser M, Lu L. Familial lymphoid interstitial pneumonia. Pediatrics 1980;65:525–528.
16. Scully RE, Mark EJ, McNeely WF, Ebeling SH, Ellender SM. Case records of the massachusetts general hospital: weekly clinicopathological exercises. N Engl J Med 1999;341:2075–2083.
17. Nogee LM, Dunbar AE, West SE, Askin F, Hamvas A, Whitsett TE. Mutation in the surfactant protein C gene associated with familial interstitial lung disease. N Engl J Med 2001;344:573–579.
18. Northway WH, Rosan R, Porter R. Pulmonary disease following respiratory therapy for hyaline membrane disease: bronchopulmonary dysplasia. N Engl J Med 1967;276:357–368.
19. Kikkawa Y, Kaibara M, Motoyaura EK, Orzalesi MM, Cook CD. Morphologic development of fetal rabbit lung and its acceleration with cortisol. Am J Pathol 1971;64:423–442.
20. Maniscalo WM, Wilson CM, Gross I, Gobran LI, Rooney SA, Warshaw JB. Development of glycogen and phospholipid metabolism in fetal and newborn rat lung. Biochem Biophys Acta 1978;530:333–346.
21. Rooney SA, Gobran LI, Chu AJ. Thyroid hormone opposes some glucocorticoid effects on glycogen content and lipid synthesis in developing fetal rat lung. Pediatr Res 1986;20:545–550.
22. Farell PM, Bourbon JR. Fetal lung surfactant lipid synthesis from glycogen during organ culture. Biochem Biophys Acta 1986;878:159–167.
23. Tyler NK, Hyde DM, Hendrickx G, Plopper CG. Cytodifferentiation of two epithelial populations of the respiratory bronchiole during fetal lung development in the Rhesus monkey. Anat Rec 1989;225:297–309.
24. Vaccaro C, Brody JS. Ultrastructure of developing alveoli: I. the role of the interstitial fibroblast. Anat Rec 1978;192:467–480.
25. Kaplan NB, Grant MM, Brody JS. The lipid interstitial cell of the pulmonary alveolus: age and species differences. Am Rev Respir Dis 1985; 132:1307–1312.
26. McGowan SE, Torday JS. The pulmonary lipofibroblast (lipid interstitial cell) and its contribution to alveolar development. Ann Rev Physiol 1997;59:43–62.
27. Burri PH. Development and growth of the human lung. In: Fishman AP, editor. Handbook of physiology. Bethesda, MD: American Physiology Society; 1985. p. 1–31.
28. Kerem E, Bentur L, England S, Reisman J, O'Brodovich H, Bryan AC, Levison H. Sequential pulmonary function measurements during the treatment of infantile chronic interstitial pneumonitis. J Pediatr 1990; 116:61–67.
29. Sharief N, Crawford OF, Dinwiddie R. Fibrosing alveolitis and desquamative interstitial pneumonitis. Pediatr Pulmonol 1994;17:359–365.
30. Mason RJ, Schwarz MI, Hunninghake GW, Musson RA. Pharmacological therapy for idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 1999;160:1771–1777.
Correspondence and requests for reprints should be addressed to Hugh O'Brodovich, M.D., FRCP (C), Professor of Paediatrics and Physiology, Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada, M5G 1X8. E-mail:

Related

No related items
American Journal of Respiratory and Critical Care Medicine
165
11

Click to see any corrections or updates and to confirm this is the authentic version of record