INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Since 1994 an endemic chronic obstructive pulmonary disease (COPD) has developed in Taiwan after a wide-spread, prolonged, and unregulated use of an allegedly body-weight- reducing vegetable, Sauropus androgynus (SA) (1, 2). Typically, the patients are young or middle-aged females with real or self-assumed weight problems, but without any previous history of respiratory ailments. The routine chest radiograph is normal, but high-resolution computed tomography (CT) shows diffuse bronchiectasis and patchy mosaic perfusion with focal attenuation. Studies of limited lung biopsy specimens from those patients have revealed bronchiolitis obliterans organizing pneumonia (BOOP) (1) or bronchiolitis obliterans (BO) (2, 3). The precise mechanism of pathogenesis is uncertain. T-cell-mediated immunity, however, has been suspected in the development of this disease (2, 3).

All conventional treatments for COPD, including steroids and bronchodilators, have been ineffective in those patients with SA-induced lung disease. A few patients have died, but many have developed protracted chronic respiratory failure. Because of the chronic debilitation and ineffective conventional treatments, single lung transplants were performed as the last resort in four patients at the National Taiwan University Hospital. The morphologic studies of the excised lungs constitute the present report.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Brief Summary of Clinical History

All four patients were female nonsmokers 34, 38, 44, and 45 of yr age. All were healthy, slightly overweight, and without respiratory ailments until the onset of the present illness. All patients drank the extract of ground leaves of SA, mixed with a variety of fruit juices daily, as recommended by the vendors, for 20 d to 2 mo. The daily doses taken and some of the major clinical data are summarized in Table 1.

Cough and shortness of breath developed gradually in all soon after the intake of SA began. The chest radiographs were all normal, but high-resolution computed tomography (CT) showed diffuse bronchiectasis and patchy mosaic perfusion with focal attenuation. Although a biopsy was not carried out, the findings were considered diagnostic of BO or BOOP (4). The shortness of breath progressed despite discontinuation of SA intake and despite steroids, bronchodilators, and other treatments for COPD. All developed protracted chronic respiratory failure (FEV₁ less than 0.6 L) (Table 1). Single lung transplants were performed as the donor lungs became available.

Protocol

All recipient lungs obtained were sliced into slices 1 to 2 cm thick and studied grossly. Multiple sections from each slice were processed routinely and embedded in paraffin. Serial sections 6 µm thick were stained with hematoxylin-eosin, Masson trichrome, PAS, and elastica Van Gieson for light microscopy. Selected sections were stained immunohistochemically, using antibodies against CD26 for B-cells, UCHL-1 for T-cells, and cytomegalovirus.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Gross Findings

Four lungs from four patients were available for study. All lungs appeared similar and were described together. The pleural surfaces were slightly dull focally but otherwise unremarkable. The upper lobes were usually well aerated, but the lower lobes were often congested and unevenly atelectatic. The cut surfaces were pink and the tissues were generally soft by palpation. Only one lung showed occasional foci of consolidation of 1 to 2 cm in size. The lobar and segmental bronchi appeared normal. The walls of some subsegmental and smaller-sized bronchi, down to 4 mm in size, appeared irregularly thickened with markedly narrowed lumens. However, no definite bronchiectasis or emphysema was grossly apparent.

Light Microscopy

The lobar and segmental bronchi were generally well preserved. In the subsegmental and bronchi down to approximately 4 mm in diameter, the epithelial lining cells were mostly unaltered but focally inflamed, necrotic, denuded, or hyperplastic. In those bronchial walls the mucous glands were often atrophic with dilated ducts (Figures 1 and 2) showing inspissated secretions. The bronchial cartilage often appeared focally necrotic, fragmented, calcified, or ossified (Figure 2). Smooth muscle bundles were prominent. Inflammatory infiltrates, including lymphocytes and other mononuclear cells, were mostly sparse but ubiquitous. Neural tissues, many with ganglion cells, were easily found in the wall. The bronchovascular bundle around those bronchi, the normally loose interstitial space, showed increased collagen deposition, which appeared deep blue in trichrome-stained sections. In general, the degree of fibrosis of the interstitial space seemed to increase as the bronchi became smaller.

View larger version (140K): [in this window] [in a new window]   Figure 1.   A bronchus approximately 5 mm in diameter (B) and accompanying pulmonary artery (P) and pulmonary vein (V); all show mild alteration of lining cells and slightly thickened walls. The small bronchiolar and alveolar structures are little altered. Hematoxylin-eosin stain; original magnification: ×40.

View larger version (141K): [in this window] [in a new window]   Figure 2.   A higher magnification of Figure 1. The fibro-obliteration of a bronchial artery (A), ossification of the bronchial cartilage (C), and atrophic dilatation of bronchial gland (G) in Figure 1 are better discerned. The bronchial lining cells (upper half of the figure) and alveoli (lower half of the figure) are almost normal. Smooth muscle cells (M) and scattered chronic inflammatory infiltrates are slightly increased. Hematoxylin-eosin stain; original magnification: ×100.

In the largest bronchi, the bronchial arteries were often only slightly dilated. However, those in bronchi 4 to 5 mm in size often showed focally narrowed, almost completely obliterated, lumens mainly because of marked intimal proliferation of smooth muscle cells and fibrosis (Figure 2).

Bronchi of 2 to 4 mm in size, contrary to the relatively mild changes observed in larger bronchi, often showed segmental, partial, or completely circumferential, necrosis of the walls, with inflammatory infiltrates, including monocytes, lymphocytes, plasma cells, and a few neutrophils and eosinophils (Figures 3 and 4). Cartilage, muscle cells, and bronchial arteries were rarely found in the necrotic walls. Some of the bronchial walls showed massive infiltration of foamy histiocytes, which appeared to be mostly confined to, or predominantly in, the bronchial submucosa (Figure 4). Those foamy histiocytes, however, occasionally extended into the bronchial lumen or through the bronchial wall and into adjacent alveoli. In those necrotic bronchi, the bronchial lumen was often filled by a mixture of mucus, inflammatory cells, necrotic debris, or granulation tissues. Some of them appeared to be completely obliterated by organized granulation and fibrous tissues (Figure 5), or dilated with fibrosis and proliferation of smooth muscle cells in the wall (Figure 6).

View larger version (155K): [in this window] [in a new window]   Figure 3.   A portion of a small bronchus of approximately 3 mm in diameter showing inflammatory cells involving the lumen and the wall (N). Foamy histiocytes (H) fill the lumen, infiltrate the bronchial wall, and spill into the adjacent alveoli. The wavy bronchioles (arrows) distal to the necrotic bronchus show irregular fibrosis and atrophy with detached mucosae. Alveoli and small pulmonary vessels are mostly normal. Hematoxylin-eosin stain; original magnification: ×40.

View larger version (148K): [in this window] [in a new window]   Figure 4.   A higher magnification of a portion of the same bronchus in Figure 3. The bronchial wall is almost completely replaced by scattered foamy histiocytes, other types of mononuclear cells, a few neutrophils, and red blood cells. The alveoli adjacent to the bronchial wall show mild reactive changes, including hyperplastic type II cells at the border of the two structures. Hematoxylin-eosin stain; original magnification: ×200.

View larger version (140K): [in this window] [in a new window]   Figure 5.   A small bronchus (B), with associated fibrotic cartilage (C), is almost completely obliterated by organizing granulation tissue. The accompanying pulmonary artery (A) shows only mild intimal and muscle wall changes. Fibrosis and mild chronic inflammation of the interstitium are, however, present. Hematoxylin-eosin stain; original magnification: ×100.

View larger version (167K): [in this window] [in a new window]   Figure 6.   A dilated small airway (Right upper field ) with flattened lining cells, chronic inflammation, fibrosis, and smooth muscle proliferation of the wall. Because cartilage and bronchial vessels are not found in the wall, it's uncertain whether this is a dilated bronchus or bronchiole. However, smaller bronchioles, pulmonary vessels, and alveoli present in the same figure appear normal. Hematoxylin-eosin stain; original magnification: ×40.

Some small bronchi less than 2 mm in diameter and a few bronchioles, distal to or closely associated with their proximal necrotic bronchi, also showed similar obstructive (Figure 7) or dilated changes. Smooth muscle cells often became hyperplastic in the fibrotic walls of dilated bronchioles. In addition, some of them were also infiltrated by lymphocytes and other inflammatory cells as in the small bronchi. Many bronchioles less than 1 mm in size, however, appeared unaltered (Figure 6). Immunohistochemical stains revealed that lymphocytes expressed mixed T- and B-phenotypes, although T-lymphocytes predominated.

View larger version (163K): [in this window] [in a new window]   Figure 7.   A terminal bronchiole is completely obliterated by loose connective tissue. Chronic inflammation and fibrosis and collection of foamy histiocytes are all mild around the bronchiole and pulmonary artery, and in adjacent alveoli. Hematoxylin-eosin stain; original magnification: ×100.

Alveoli were generally well preserved (Figures 6 and 7). Rarely, mucus and macrophages from obstructed bronchioles extended into the adjacent alveoli with mild interstitial inflammation and fibrosis. In one lung with gross evidence of consolidation, focal cytomegalovirus pneumonitis was confirmed by immunostaining (figure not shown).

The pulmonary arteries and veins had only focal mild to moderate increase in the thickness of intima and muscle layer proper (Figure 1). The pleura and lobular septum showed focal mild fibrosis. Lymphatic channels were dilated focally. There was no evidence of vasculitis or granuloma formation in all sections studied.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

With the whole lung available for study, we found that the major light microscopic (LM) changes in the lungs of patients with SA-related COPD were segmental necrosis, fibrosis, and obliteration of small bronchi 2 to 4 mm in size. The previously reported obstructive or constrictive bronchiolitis (BO) (1) was also present, but the extension of intraluminal exudate into the alveolar ducts and alveoli (BOOP) (2) was rare. These bronchiolar changes were most likely a distal extension of the major changes in the small bronchi. The larger bronchi proximal, and smaller bronchioles and alveoli distal, to the necrotic segments were either mildly altered or normal.

The segmental destruction of many small bronchi with relative sparing of larger bronchi, most small bronchioles, and alveoli was different from the usual LM findings of COPD and also many other types of obstructive bronchiolitis that have been described (5, 6). These findings, however, could explain why the patients had severe respiratory signs and symptoms yet the chest radiographs remained normal (1), and why the excised lungs had deceptively normal gross appearances. Also, patients with most of their small bronchi obliterated by necrosis or fibrosis would be refractory to conventional treatments for COPD.

The precise mechanism that led to this unusual combination of pathologic changes is uncertain. However, the wide-spread atrophy, degeneration, and fibrosis of bronchial glands and cartilage in the large bronchial walls were most consistent with chronic ischemia. The sclerotic narrowing or obliteration of bronchial arteries in the large bronchi might be responsible for those ischemic changes. They further might cause poor blood perfusion and ischemic necrosis in the segments of bronchi at the water-shed zone between the systemic and pulmonary circulation. Pulmonary arteries and veins were only slightly sclerotic and probably played less important roles in the pathogenesis.

The absence of any inflammatory infiltrates in and around the bronchial arteries supported the idea that the vascular changes in large bronchi preceded the necrotic changes of small bronchi 2 to 4 mm in size. The degree of ischemia, necrosis, obstruction, and secondary infection in these segments, in turn, influenced the development of varied cystic dilatation, fibrotic narrowing, or obliteration of some bronchioles and alveoli further distally. The steroid therapy could have contributed to the development of cytomegalovirus pneumonitis.

The focal collections of submucosal foamy histiocytes that spilled focally into the bronchial lumen or through the bronchial wall into the adjacent alveoli were intriguing. Although a remote submucosal hemorrhage was a possibility, no associated hemosiderin pigments were apparent. They were reminiscent of the collections of foamy histiocytes in atherosclerosis.

Contrary to the CT findings, bronchiectasis was not obvious, grossly, and many dilated bronchioles were not acutely inflamed, as shown by LM. Bronchiectasis, in the conventional pathologic sense, did not exist in these resected lungs. However, bronchioles distal to some incompletely obstructed necrotic bronchi were probably dilated in situ in the chest cavity because of air trapping. If the proximal bronchi were completely obstructed, some collapsed alveoli could also pull adjacent bronchiolar walls ectatic. The alternatively hyperinflated or collapsed alveoli, as seen in the gross lungs, and similar to that in BO (4), would explain the CT findings of mosaic perfusion of lung parenchyma with patchy low attenuation.

The agent or agents responsible for the pathophysiologic events in these lungs were not certain. Viral and Mycoplasma infections, exposures to toxins, fumes, or insecticides, drug reactions, and many more, have been considered and excluded (1, 2). The only known exposure common to all patients in this group was a prolonged daily ingestion of the ground fresh leaves, or their filtered juice, of a plant that allegedly can reduce body weight (1, 2). The shrub, Sauropus androgynus (Sauropus albicans, asin-asin, or Chekor manis), is a member of the family Euphorbiacease, which allegedly contains about 580 mg of papaverine per 100 g (7, 8). Respiratory problems from ingestion of SA, however, have not been reported until this episode in Taiwan.

At least several hundred SA-related cases of COPD have been documented in Taiwan (1, 9, 10). Although the modes and dosages of SA ingestion are quite varied and uncontrolled among all patients, the development of COPD appears to be dose-dependent (9). In some patients, respiratory symptoms have regressed spontaneously after an early cessation of the ingestion of SA (9). Subclinical SA-induced lung changes also exist and can be demonstrated by ^99mTc-DTPA radioaerosol inhalation scan in some (10). Many attempts to reproduce this disease in small rodents, including guinea pigs, rats, and mice, however, have not been successful.

It may be argued that the bronchial arterial changes were secondary to obstructive bronchitis. However, the bronchial arterial changes were not present in the necrotic segment, but in the wall of the larger, more proximal, and relatively normal-appearing, segment of the bronchi. This relationship is somewhat similar to that of the atherosclerotic obliteration of proximal femoral arteries causing ischemic necrosis of peripheral toes. In our patients, because of the dual blood supply, ischemic necrosis occurred only in the small bronchi, which depended entirely on bronchial arterial supply. The smallest bronchi, bronchioles, and alveoli were preserved mostly because of the relatively unaltered, or compensatory, pulmonary arterial perfusion.

Also, the histology of SA-induced necrotizing bronchitis may superficially resemble that of the bronchocentric granulomatosis. The latter, however, is a clinically benign disease that usually responds dramatically to steroid therapy.

In summary, we have documented that necrosis and fibrotic obstruction of bronchi 2 to 4 mm in size are the main changes in patients who developed COPD after prolonged use of SA. The agent responsible for these pathologic changes has not been definitively identified. However, the segmental necrosis of bronchi occurred in the watershed region of the bronchial and pulmonary circulation and could be the results of a prolonged, compromised perfusion from fibrosclerotic bronchial arteries. Foci of secondary obstructive or cystic and ectatic bronchiolitis occurred distally.

The combined pathologic findings of segmental destruction of many small bronchi with relative sparing of large bronchi, most small bronchioles, and alveoli, were unique and different from those found in usual COPD. Because conventional therapies for COPD have been ineffective, these pathologic findings probably should warrant consideration in the planning of new therapeutic strategies such as creating collateral passages between the large bronchi and peripheral airways in these patients.