Diagnosis: Pulmonary thrombotic angiopathy and fat embolism of sickle cell disease, mild bronchitis (clinical acute chest syndrome)

Comment: This patient developed respiratory failure and pulmonary opacities while in the hospital for pain crisis. Although she did not have a history of lung problems, she had evidence at autopsy of old and recent pulmonary thrombotic angiopathy: old and organizing parenchymal scars and organized thrombi, as well as widespread acute microvascular distension by sickle RBCs, focal alveolar wall necrosis, and alveolar edema. Fat emboli were also present. Although there was no evidence of bacterial pneumonia, the upper respiratory (? viral) infection may have predisposed to the final acute event.

Two supplementary cases illustrate other features of lung disease in sickle cell disease: supplementary case #1 and supplementary case #2.

Lung Abnormalities in Sickle Cell Disease (SCD)

Introduction: In the US, around 85% of persons with SCD now live past age 20, but life expectancy is reduced by about 10-15 y compared to that of the healthy African-American. Pulmonary disease occurs in up to 50% of persons, 20-80% of them have recurrences, and about 25% of them die of lung disease [1]. Two syndromes have been described: acute chest syndrome (ACS) attributed to infection or infarction, or a combination, and chronic lung disease, marked by the development of permanent loss of function that progresses to pulmonary hypertension [2].

Acute chest syndrome (ACS)

Clinical features: ACS is defined as the appearance in a patient with SCD of a new radiographic opacity, or a new perfusion defect on a radioisotope scan. Accompanying symptoms (in a study of 1,722 episodes) may include fever, cough (sometimes productive), chest pain, dyspnea, chills, wheezing, or hemoptysis. Incidence varies with type of SCD (SS > sickle cell-betao thalassemia > SC > sickle cell-beta+ thalassemia) [3]. Other risk factors include high hematocrit and WBC count, and low Hb F concentration [2]. The incidence is high in children 2 to 4 y old and less in adults [3]. Fever and cough are present more often in young children <2 years old and indicate a higher frequency of infection at that age. Pain in the extremities precedes ACS more frequently in adults than in children. In any individual case, recurrent episodes of ACS tend to be accompanied by the same symptoms [4].

Course: In the study of 1,722 episodes of ACS (above), hemoglobin fell from baseline (ave of 0.7 g/dl), and WBC rose by an average of 69%. Hypoxemia was present. Bacteremia occurred in children <2 y old in 14% but in patients 10 or more years old, in only 1.8%. Episodes occurred more frequently in winter months in all age groups. (It should be noted that the 3 patients presented here all died shortly after the onset of an upper respiratory tract infection in the winter.) The length of hospitalization in children was less than that in adults (5.4 vs 9 da). In adults, hemoptysis, productive cough, dyspnea, and high respiratory rate predicted a longer course. Death rate per ACS admission was 1.8% (children <20 y 1.1%, adults 4.3%). Multilobar disease was present in most fatal cases. In those who died, respiratory failure developed within 48 hr of admission, sepsis was common, and vaso-occlusive crises in extremities were present. At autopsy massive fat embolism was found in 9/16 [4].

Goals for Treatment of ACS [1]

Chronic sickle cell lung disease

This entity has been defined as permanent lung disease that is restrictive in nature. It has been divided into 4 stages [6]. Stages 1 to 4 show progressive restriction, and stage 4 is characterized by pulmonary hypertension. Although staging criteria describe diffuse fibrosis in all lung lobes on plain films, diffuse fibrosis has not been described in CT studies [7]. Also, autopsy studies show little fibrosis [2]. In a longitudinal study of 28 patients, the main predictor of chronic disease was number of episodes of ACS. Mean time to death after diagnosis of chronic disease was 5 years [6].

Radiographic changes

Acute chest syndrome: In the study of 1,722 episodes of ACS, children tended to have isolated upper or middle lobe disease, and adults tended to have multilobe involvement and pleural effusion [4]. Study of thin (3 mm) CT scans from 10 patients, ages 3 to 11, within the first 48 hr of admission for ACS showed peripheral microvascular hypoperfusion, areas of ground glass opacity, or dense consolidation. The ground glass opacities showed no lobar preferences. A follow-up CT in 3 was normal in 2 and improved in 1. The CT changes paralleled the clinical severity of the disease better than the opacities seen on chest radiographs [8].

Chronic lung disease: Thin (1-1.5 mm) CT scans of 29 patients, 5-54 y old (mean 22 y), with 1 to >10 (median 6) previous episodes of ACS were taken more than 1 month after the last episode. Scans were scored 0-3 for abnormalities. The main features included parenchymal bands, interlobular septal thickening, pleural tags, dilated lobules, and architectural distortion. Of the total, 59% had few changes (grade 0-1) (group A), and 41% had more changes (group B). The grade correlated with the number of previous ACS events but not with the results of PFTs. PFTs were abnormal in 11 of 14 in group A: a decrease in DLCO and FVC was frequent. In group B patients, 8 of 10 had abnormal PFTs, which were mainly restrictive. No patient had diffuse interstitial pulmonary fibrosis or honeycombing [7]. If the diagnosis of chronic lung disease requires abnormal PFTs [6], then 5 of the 24 patients tested did not qualify.

Histologic changes

Introduction: The terms "infection" and "infarction" have different connotations in the patient with SCD than in other persons. Typical bacterial pneumonia is rare, at least in the adult, but viral (or mycoplasmal or chlamydial) infection may promote sickle RBC-endothelial interactions and predispose to ACS with little or no histologic evidence of pneumonia (see pathogenesis). Infarction usually implies embolization to segmental arteries, a shadow that progresses to a linear scar on plain film, and compromise to the venous flow. The term "thrombotic angiopathy of SCD" is introduced here to refer to the unusual changes that occur in the microvasculature of patients with SCD. Fat embolism is described separately although it may be an integral part of the vascular obliteration that occurs.

Pulmonary thrombotic angiopathy of SCD: The lungs of patients dying of respiratory failure without bacterial pneumonia show alveolar edema and widespread distension of the microvasculature by sickle RBCs. On closer inspection of congested areas, alveolar wall necrosis that resembles a capillaritis may be present [9]. It consists of alveolar wall destruction, fibrinous exudate, and PMNs in the wall and adjacent air spaces. Its cause appears to be ischemia.

Here and there, widening of alveolar walls by granulation tissue and interstitial macrophages with light-brown pigment constitutes organization. Increased numbers of the same macrophages can be found with PMNs in adjacent alveoli. This lesion, probably ischemic in nature and perhaps continuing between acute episodes, may progress to scar, which can be in the center of the acinus or subpleurally in a wedge-shape. Metaplastic epithelium may line the scarred walls, and mucus can fill the involved air spaces. Pleural adhesions are a common result of the pulmonary ischemia.

Larger vessels are also involved. Patchy intimal fibrosis of venules, arterioles, and muscular arteries can be highlighted by an elastic stain [10]. In arteries, the thickening may be eccentric (suggesting organized thrombi) or concentric (possibly implying another mechanism, perhaps cytokine mediated) [11]. With progression, acute and organizing thrombi are found in medium-sized and large muscular arteries usually without evidence of large infarcts [10, 12]. Plexiform changes have been described before in one case [11], and plexogenic angiopathy has been observed in non-SCD patients with thromboembolic pulmonary hypertension [13]. Thrombi have been described in a main pulmonary artery in 2 cases of SCD [14]. Similar proximal thrombi, ascribed to low flow, have been described in other patients with pulmonary hypertension without sickling [15].

The vascular pattern described above differs from the scattered capillary fibrin thrombi of DIC, which is usually accompanied by diffuse alveolar damage. The pattern also differs from the focal occlusion of muscular arteries by thromboemboli, which may be accompanied by segmental pulmonary hemorrhage and edema or infarcts, which are usually larger than those described in SCD. Most reports of autopsied patients with SCD, however, have not included a statement about examination of the leg veins [9,12,14,16], nor were they described in any of the 3 cases just presented. Although it is unlikely that thromboemboli are the cause of the changes in SCD, this possibility needs to be considered and excluded.

Summary of Histologic Changes of Pulmonary Thrombotic Angiopathy of SCD

Fat/bone marrow embolism: Fat emboli along with fragments of necrotic bone marrow may lodge in the lung. Fat droplets that traverse the capillary bed may also be found in the systemic circulation [16]. None was found in the brain or kidney of the present patient. The fat in the lung can be taken up by alveolar macrophages. These lipid-laden macrophages may then be identified in BAL fluid by the oil red O stain. In these cells, the fat must be distinguished from lipochrome pigment. Fat globules are larger and redder than the granular, brownish, lipochrome pigment, but examination of oil red O-stained samples before and after treatment with lipid solvents helps to distinguish lipid from lipochrome. Lipid solvents remove the oil red O-positive lipid but not the lipochrome [17]. The role of fat emboli in progression of the vascular disease is unknown. Embolization can be associated with sudden death [16].

Pathogenetic mechanisms involved in the evolution of these changes are described below (Pathogenesis).

Cardiac findings: Because ACS can resemble congestive heart failure, hearts of 52 patients dying of sickle cell disease were studied for signs of a cardiomyopathy. No infarcts or coronary thromboses were noted. All the changes could be related to chronic anemia rather than to a specific cardiomyopathy of SCD [18]. Others have described myocardial fibrosis and infarcts in patients with chronic sickle cell lung disease [6]. Whether these are due to thrombotic angiopathy or secondary to hypoxia, as appeared to be the cause in the case under discussion, is unknown.


Sickling increases the viscosity of blood and predisposes to stasis especially in the pulmonary artery, where factors such as low pH and oxygen tension (especially in anemia) promote it. Pulmonary complications, such as thrombosis, ischemia, necrosis, and fibrosis, depend on the percentage of hemoglobin S, presence of fever and leukocytosis, exercise, and amount of physiologic or pathologic shunting [19]. Besides sickling, activation of the coagulation system (hypercoagulable state) and RBC-endothelial interactions contribute to vascular damage and to the repeated episodes of acute chest syndrome. Infections (bacteria, viruses, chlamydia, mycoplasma), which are difficult to distinguish from ischemic episodes, and bone marrow emboli also play a role. Multiple episodes of ACS are a risk factor for chronic lung disease. The possible contribution of transfusion and rehydration to acute capillary leak and diffuse pulmonary edema is suspected [2]. Some of these factors are explored below.

1. RBC-endothelial interaction: RBCs of patients with sickle cell anemia have a survival time of 10-30 da (normal 120 da). One factor contributing to early loss is the coating of the older, denser cells (more frequent in SS blood than in AA blood) by natural antibody found in normal serum. This antibody (anti-alpha-galactosyl IgG) binds to the RBC via the Fab portion of the molecule and leaves the Fc portion available to interact with macrophages and be destroyed or to interact with virally-infected endothelial cells [20]. The mechanism by which viral infection predisposes to vaso-occlusive events was studied in an in vitro model of RBC adherence to cultures of human umbilical vein endothelial cells. When endothelial cells were infected with HSV-1 (one of many types of virus that infect endothelium), washed RBCs from SS patients adhered 1.8 times more frequently than those from AA persons. The phenomenon was caused by the appearance of Fc receptors on the endothelium, which allows it to react with IgG on sickle RBCs [21].

Other factors are also important for endothelial adherence of sickle RBCs. In vitro studies have shown that fibrinogen in acute phase plasma increases the adherence of sickle RBCs to normal endothelium [22].

In a model to define the localization of RBC adhesion in SCD, washed RBCs were perfused through the mesocecal microvasculature of the rat. Cell adhesion was greatest (SS>>AA) in the postcapillary venule with concentrations that correlated inversely with the venular diameter [23].

Another measure of vascular damage is the increase in number of circulating endothelial cells that can be identified in the buffy coat. Cells were identified by an immunostain in 18 patients with SCD during a pain crisis and between crises, and in 14 normal persons. During the crisis, an average of 23 endothelial cells/ml blood was found and between crises, 13/ml. The normal controls, including patients with hemolytic anemia, had 2-3/ml. The cells were mainly CD36-positive (i.e., derived from the microvasculature) and expressed the activation markers ICAM-1, VCAM-1, E-selectin and P-selectin [24]. In another study, endothelin-1 was found to be elevated in plasma just before and during ACS [25].

Finally, it has been speculated that some of the vascular intimal thickening may be caused by cytokines released by RBC-endothelial interactions or by platelet adhesion and release of platelet derived growth factor. Organizing thrombi may not be the sole cause [11].

2. Pulmonary fat embolism is a complication that has been described in patients with serious traumatic fractures. Diagnosis is based on a triad of clinical findings (Table), which may be confirmed by finding oil red O-positive alveolar macrophages or PMNs (>5% of lavage cells scanned at 1000 x) in BAL fluid obtained from areas of lung showing radiographic opacities. Patients without the syndrome, patients with ARDS, and normal controls had 5% or fewer oil red O-positive cells [26].

Table: Clinical Manifestations of Traumatic Fat Embolism [26]

In patients with SCD, fat embolism is believed to contribute to the development of ACS. It is a result of hematogenous dissemination of necrotic marrow from bone infarcts that occur during crisis. It has been documented histologically at autopsy in 13% of 72 patients in one study [9]. Death from fat embolism was described in one report [16]. BAL has also been used to diagnose fat embolism in patients with SCD. In patients, ages 2 to 23 y, 12 of 27 had increased numbers of oil red O-positive macrophages. These patients differed from those with few lipid-laden macrophages in having significantly more pain crises, chest pain, and neurologic symptoms than those without. Also, they had lower hemoglobin and platelet levels and more nucleated RBCs than those without. The hospital course was longer (13 vs 7 da). The authors concluded that fat embolism is a distinct cause of ACS [27]. Godeau, et al. did a similar study in adults (mean age 26 y). They found evidence of fat embolism in 12 of 20 patients, 11 of whom had definite or probable bone marrow infarcts. In contrast to the previous report, they found no clinical or radiographic differences between those with or without fat embolism. Thus, they could not confirm that it is an important etiologic factor in ACS [17]. Reports by others indicate that macrophages containing oil red O-staining droplets are not specific for fat embolism, and care should be taken in the interpretation of the results [28,29].

Nevertheless, fat embolism does occur and may be an important factor contributing to ACS because the fat and cell membranes undergo hydrolysis to liberate toxic free fatty acids. In this situation, the enzyme phospholipase A2 (PLA2) may be involved in producing the free fatty acids, such as oleic and arachidonic acid. The latter can give rise to thromboxane, leukotrienes, and prostaglandins. It has been found that patients with SCD have markedly elevated values of PLA2. Those with ACS (n=20) had PLA2 values averaging 336 ng/ml plasma, whereas those with vaso-occlusive crisis (n=10), those in a steady state (n=11), and normal controls without sickle disease (n=19) had values of 24, 10, and 3.1, respectively. A group with pneumonia but without sickle disease (n=11) had a mean of 69 ng/ml [30]. The level of PLA2 correlated with severity of the chest syndrome: it was higher in patients with lower PaO2, with increased (A-a) O2 gradients, and in those requiring transfusion. In one patient studied during hospitalization for vaso-occlusive crisis, PLA2 rose to high levels 2 days preceding ACS. Thus, it may be a useful marker for impending ACS. TNF and Il-1 upregulate PLA2, and both may be elevated in patients with sickle cell disease. PLA2 is also increased in patients with ARDS in proportion to its severity [30].

New treatments

Incentive spirometry: Infarcts in bones of the thorax can be associated with hypoventilation because of pain. In a prospective, randomized trial including 29 patients, ages 8 to 21, during 38 episodes of acute chest pain or back pain above the diaphragm, the incidence of bone infarcts was examined by bone scan two or more days after admission. During the same episode, the value of incentive spirometry (10 maximal inspirations at 2-hour intervals from 8 am to 10 pm) to decrease the incidence of pulmonary opacities was also studied. Chest radiographs were obtained on admission and three or more days later. Of the 15 with bone infarcts, none of 7 in the spirometry group developed ACS compared to 5 of 8 in the nonspirometry group (P=0.025) [31]. Thus, this maneuver may prevent pulmonary opacities by alleviating atelectasis and consequent sludging of blood in hypoventilated lung.

Hydroxyurea: Based on the hypothesis that an increased amount of fetal hemoglobin will diminish intravascular sickling and reduce the number of painful crises, a trial of hydroxyurea, a cytotoxic drug that stimulates production of fetal hemoglobin, was undertaken. A double-blind study of 299 patients with 3 or more painful crises per year showed that the 152 patients who received hydroxyurea had fewer painful crises (2.5 vs 4.5/y) and fewer episodes of ACS (25 vs 51). The long-term safety of the treatment is still unknown, and monitoring and patient compliance are important [32].

Bone marrow transplantation: This treatment is an expensive but potential cure for severe disease. Successful transplants were performed in 12 patients between the ages of 11 mo and 23 y (median 4 y), although several severe complications occurred. Patients were alive at follow-up from 9 to 51 mo. Vaso-occlusive crises and hemolysis were abolished [33].

Link to other sickle cell disease sites.


1. Dreyer Z. Chest infections and syndromes in sickle cell disease of childhood. Sem Respir Infect 1996; 11:163-172.

2. Weil J, Castro O, Malik A, Rodgers G, Bonds D, Jacobs T. Pathogenesis of lung disease in sickle hemoglobinopathies. Am Rev Respir Dis 1993; 148:249-256.

3. Castro O, Brambilla D, Thorington B, Reindorf C, Scott R, Gillette P, Vera J, et al. The acute chest syndrome in sickle cell disease: incidence and risk factors. Blood 1994; 84:643-649.

4. Vichinsky E, Styles L, Colangelo L, Wright E, Castro O, Nickerson B, et al. Acute chest syndrome in sickle cell disease: clinical presentation and course. Blood 1997; 89:1787-1792.

5. Emre U, Miller S, Rao S, Rao M. Alveolar-arterial oxygen gradient in acute chest syndrome of sickle cell disease. J Pediatr 1993; 123:272-275.

6. Powars D, Weidman J, Odom-Maryon T, Niland J, Johnson C. Sickle cell chronic lung disease: prior morbidity and the risk of pulmonary failure. Medicine 1988; 67:66-76.

7. Aquino S, Gamsu G, Fahy J, Claster S, Embury S, Mentzer W, Vichinsky E. Chronic pulmonary disorders in sickle cell disease: findings at thin-section CT. Radiology 1994; 193:807-811.

8. Bhalla M, Abboud M, McLoud T, Shepard J-A, Munden M, Jackson S, Beaty J, et al. Acute chest syndrome in sickle cell disease: CT evidence of microvascular occlusion. Radiology 1993; 187:45-49.

9. Haupt H, Moore G, Bauer T, Hutchins G. The lung in sickle cell disease. Chest 1982; 81:332-337.

10. Collins F, Orringer E. Pulmonary hypertension and cor pulmonale in the sickle hemoglobinopathies. Am J Med 1982; 73:814-821.

11. Case Records of the Massachusetts General Hospital. N Engl J Med 1983; 309:1627-1636.

12. Oppenheimer E, Esterly J. Pulmonary changes in sickle cell disease. Am Rev Respir Dis 1971; 103:858-859.

13. Moser K, Bloor C. Pulmonary vascular lesions occurring in patients with chronic major vessel thromboembolic pulmonary hypertension. Chest 1993; 103:685-692.

14. Rubler S, Fleischer R. Sickle cell states and cardiomyopathy. Sudden death due to pulmonary thrombosis and infarction. Am J Cardiol 1967; 19:867-873.

15. Moser K, Fedullo P, Finkbeiner W, Golden J. Do patients with primary pulmonary hypertension develop extensive central thrombi? Circulation 1995; 91:741-745.

16. Shapiro M, Hayes J. Fat embolism in sickle cell disease. Report of a case with brief review of the literature. Arch Intern Med 1984; 144:181-182.

17. Godeau B, Schaeffer A, Bachir D, Fleury-Feith J, Galacteros F, Verra F, Escudier E, et al. Bronchoalveolar lavage in adult sickle cell patients with acute chest syndrome: value for diagnostic assessment of fat embolism. Am J Respir Crit Care Med 1996; 153:1691-1696.

18. Gerry Jr J, Bulkley B, Hutchins G. Clinicopathologic analysis of cardiac dysfunction in 52 patients with sickle cell anemia. Am J Cardiol 1978; 42:211-216.

19. Moser K, Shea J. The relationship between pulmonary infarction, cor pulmonale and the sickle states. Am J Med 1957; 22:561-579.

20. Galili U, Clark M, Shohet S. Excessive binding of natural anti-alpha-galactosyl immunoglobin G to sickle erythrocytes may contribute to extravascular cell destruction. J Clin Invest 1986; 77:27-33.

21. Hebbel R, Visser M, Goodman J, Jacob H, Vercellotti G. Potentiated adherence of sickle erythrocytes to endothelium infected by virus. J Clin Invest 1987; 80:1503-1506.

22. Hebbel R, Moldow C, Steinberg M. Modulation of erythrocyte-endothelial interactions and the vasocclusive severity of sickling disorders. Blood 1981; 58:947-952.

23. Kaul D, Fabry M, Nagel R. Microvascular sites and characteristics of sickle cell adhesion to vascular endothelium in shear flow conditions: pathophysiological implications. Proc Natl Acad Sci USA 1989; 86:3356-3360.

24. Solovey A, Lin Y, Browne P, Choong S, Wayner E, Hebbel R. Circulating activated endothelial cells in sickle cell anemia. N Engl J Med 1997; 337:1584-1590.

25. Hammerman S, Kourembanas S, Conca T, Tucci M, Brauer M, Farber H. Endothelin-1 production during the acute chest syndrome in sickle cell disease. Am J Respir Crit Care Med 1997; 156:280-285.

26. Chastre J, Fagon J-Y, Soler P, Fichelle A, Dombret M-C, Huten D, Hance A, et al. Bronchoalveolar lavage for rapid diagnosis of the fat embolism syndrome in trauma patients. Ann Intern Med 1990; 113:583-588.

27. Vichinsky E, Williams R, Das M, Earles A, Lewis N, Adler A, McQuitty J. Pulmonary fat embolism: a distinct cause of severe acute chest syndrome in sickle cell anemia. Blood 1994; 83:3107-3112.

28. Vedrinne J, Guillaume C, Gagnieu M, Gratadour P, Fleuret C, Motin J. Bronchoalveolar lavage in trauma patients for diagnosis of fat embolism syndrome. Chest 1992; 102:1323-1327.

29. Stanley J, Hanson R, Hicklin G, Glazier Jr A, Ervanian A, Jadali M. Specificity of bronchoalveolar lavage for the diagnosis of fat embolism syndrome. Am Surgeon 1994; 60:537-541.

30. Styles L, Schalkwijk C, Aarsman A, Vichinsky E, Lubin B, Kuypers F. Phospholipase A2 levels in acute chest syndrome of sickle cell disease. Blood 1996; 87:2573-2578.

31. Bellet P, Kalinyak K, Shukla R, Gelfand M, Rucknagel D. Incentive spirometry to prevent acute pulmonary complications in sickle cell diseases. N Engl J Med 1995; 333:699-703.

32. Charache S, Terrin M, Moore R, Dover G, Barton F, Eckert S, McMahon R, et al. Effect of hydroxyurea on the frequency of painful crises in sickle cell anemia. N Engl J Med 1995; 332:1317-1322.

33. Vermylen C, Cornu G, Philippe M, Ninane J, Borja A, Latinne D, Ferrant A, et al. Bone marrow transplantation in sickle cell anaemia. Arch Dis Child 1991; 66:1195-1198.

Clinical summary

Comments: mw6825@itsa.ucsf.edu

Table of Contents

Copyright 1998 by Martha L. Warnock. All rights reserved.

Last revised 3/14/98