J Clin Pathol 1985;38:659-664

Vascular occlusion and infarction in sickle cell crisisand the sickle chest syndromeNA ATHANASOU,* C HATTON,t J O'D McGEE* DJ WEATHERALLt

From the University of Oxford, Nuffield Departments of *Pathology and tClinical Medicine, John RadcliffeHospital, Oxford

SUMMARY A young adult with homozygous sickle cell anaemia (Hb SS) suffered a fatal sickle cellcrisis complicated by the sickle chest syndrome. At necropsy multiple large infarcts of the lung,bone marrow, and pituitary gland were found. The large majority of pulmonary infarcts were notassociated with either gross or microscopic vaso-occulusion. These findings are discussed andcorrelated with past and current opinions of sickle cell crisis and the sickle chest syndrome.

Sickle cell crisis refers to any new syndrome thatdevelops rapidly in patients with sickle cell diseaseon the basis of the inherited abnormality and notdue to any other cause.' The term sickle cell crisis isprincipally a clinical term embracing a number ofpathological mechanisms including vaso-occlusion,bone marrow aplasia, and red cell sequestration. Inhomozygous sickle cell anaemia (Hb SS) painfulvaso-occlusive crises are the most common cause ofmorbidity and mortality in children and adults2 and,with the sickle chest syndrome (acute pulmonaryinvolvement without demonstrable bacterial infec-tion), account for more than 90% of hospital admis-sions.3Of primary importance in initiating vascular

occlusion is the decreased deformability of the sickleerythrocytes themselves. These increase blood vis-cosity, promoting stasis and augmenting localhypoxia and acidosis, which, in turn, leads to furthersickling. Other factors, however, including vascularspasm and the coagulation system are also thoughtto be involved in the pathogenesis of vaso-occlusion.' I Infection is believed to be one of theprecipitating factors of sickle cell crisis,5 but whetherthe acute chest syndrome is due to infection6 orvaso-occlusion7 or both is unclear.We report the clinical and pathological findings of

a patient with homozygous sickle cell anaemia whowas admitted with symptoms of a painful vaso-occlusive crisis and who deteriorated suddenly,showing features of the sickle chest syndrome.

Accepted for publication 14 February 1985

Case report

A 19 year old woman was admitted complaining ofgeneralised bone pain, which was particularly severein the chest, back, and limbs. She admitted to havinghad a cough producing yellow sputum over the pre-ceding two weeks but had no dyspnoea, orthopnoea,or haemoptysis. She had been born to parents ofWest Indian origin and had been diagnosed as hav-ing homozygous sickle cell disease when aged 21/2years. During her life she had averaged one or twoadmissions a year for painful crises, two of whichwere associated with diffuse lung infiltrates on chestradiograph. When she was 17 years old she had ahepatic sequestration crisis.On examination, she was obviously in severe pain

and feverish (38.2CC); her pulse rate was 110beats/min sinus rhythm, jugular venous pressure wasnot raised, and blood pressure was 130/90 mm Hg.There was no obvious jaundice, cyanosis, oroedema; she had a systolic murmur heard at the leftsternal edge and a loud pulmonary second sound.Chest examination was unremarkable. The liver waspalpable two finger breadths below the right costalmargin. Central and peripheral nervous system werenormal. Results of investigations on admission wereas follows: haemoglobin concentration 7-2 g/dl;white cell count 15 x 109/1; platelet count 582 x104/1; reticulocyte count 13 2%; coagulation screennormal; total bilirubin 31 mmolll; aspartate trans-aminase 96 IU/I; 'y-glutamyl transaminase 8 IU/I;alkaline phosphatase 271 IU/l. An electrocardio-gram showed sinus tachycardia; chest radiographshowed a mildly enlarged heart with clear lungfields. Blood and sputum cultures taken on admis-

659

on March 23, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.6.659 on 1 June 1985. Dow

nloaded from

Athanasou, Hatton, McGee, Weatherall

q

Al.

Fig. 1 Multiple lung infarcts. Irregular dark haemorrhagicareas are seen on both the cut and pleural surfaces. Nothrombus is seen in the major pulmonary vessels.

sion and during her stay in hospital were sterile andshowed no appreciable growth. Pneumococcal anti-gens were not detected.On admission she was given pethidine, intraven-

ous hydration, and antibiotics (erythromycin

250 mg three times daily and ampicillin 500 mgintravenously four times daily). The following daythe fever had not settled and she continued torequire intravenous pethidine, 75 mg every 2 h. Herblood count showed the following: haemoglobinconcentration 6.7 g/dl; white cell count 23 x 109/1;platelet count 376 x 109/l; and reticulocyte count8.8%. Twenty four hours later she became morebreathless. She was feverish with a temperature of39.8CC. Chest radiograph showed widespreadopacities. Oxygen was given in response to herdeteriorating clinical condition. Her haemoglobinconcentration was 4-9 g/dl and her white cell count28 x 109/1. She had a sudden cardiac arrest, fromwhich she was promptly resuscitated. Despite con-tinuing resuscitative measures, including bloodtransfusion, inotropic support, and mechanical ven-tilation, her condition deteriorated and she died twodays later.

NECROPSY FINDINGSThe body was that of a young, well nourished Negrowoman. No skin ulcers, skeletal, or other obviousexternal abnormalities were noted. The lungsshowed the most striking gross changes (Fig. 1).They were heavy (555 g and 545 g) and firm, andthere were multiple irregular haemorrhagic areasscattered throughout the parenchyma of both lungs.These ranged from 0-5-4-5 cm maximum dimensionand a few extended to the pleural surface. Examina-tion of the pulmonary vessels of both lungs showedno large thrombi in the pulmonary trunk, main pul-monary arteries, or immediate lobar branches.Thromboemboli, which were firm, granular, and fri-able, were present in four small conducting arteries(0-5-0-7 cm external diameter), two in each lung.These thromboemboli corresponded to the darkerareas, but the rest of the dark areas were not related

<'4) Fig. 2 Two small discretehaemorrhagic pulmonary infarctsnear the hilum ofthe lung separatedby an area of uninfarcted lung. Bar=100 pin.

660

.2F

JZ.7

on March 23, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.6.659 on 1 June 1985. Dow

nloaded from

Sickle cell crisis and the chest syndrome* . I-

H------

6.

*., '4p.

4- -'

4 * 4J

"V.

% . ..

t

Fig. 3 Margin ofsmall pulmonary infarct (top). Alveoliare filled with blood and there is necrosis ofinteralveolarsepta. At the margin with uninfarcted lung, one ofthe alveolishows hyaline membrane formation (arrow). Bar =100 ,u.

to grossly thrombosed vessels. The pulmonary veinswere normal. The heart showed left ventricularhypertrophy (ventricular weight with septum229 g); the right ventricle weighed 52 g. Themyocardium appeared normal and the major coro-nary vessels were all widely patent. No thrombi andlittle evidence of atheroma was seen in the aorta orother arteries examined. No deep vein thrombosiswas noted in the femoral or calf veins and both theportal and systemic venous circulation appearednormal. The spleen (22 g) was small (2 x 1-7 x1-2 cm), dark, and firm; there were no enlarged

lymph nodes. The stomach and small and large* intestine appeared congested. The liver (1937 g)4**, * also appeared congested and firm. The gall bladder

and biliary tree looked normal and there were no~gallstones. Both kidneys were of normal size and

showed old capsular infarcts. The cortex was wide-ned and prominent and, like the medulla, appearedcongested. The brain showed flattened gyri and nar-row sulci. The cut surface showed numerous dilated

-, * small and large blood vessels but no obvious infarc-tion. The pituitary gland appeared normal. Therewas extension of the red marrow down the femoralshaft and old bony infarcts were noted in the headand shaft of the femur, several of the vertebrae, andtne sternum.

HISTOLOGICAL EXAMINATIONNumerous recent haemorrhagic pulmonary infarctscorresponded to the dark areas seen on macroscopicexamination (Figs. 2 and 3). There was intensearteriolar and capillary congestion; the alveoli werefilled with blood and there was necrosis of inter-alveolar septa. There was a striking absence ofidentifiable thrombosis given the extent of pulmo-nary infarction. No old, organised thrombi or scarsfrom previous episodes of pulmonary infarctionwere noted and there was no evidence of pulmonaryhypertension. No fat or bone marrow emboli werepresent. Hyaline membrane formation was seen inalveoli, alveolar duct, and respiratory bronchioles ofuninfarcted lung and was presumably related tomechanical ventilation and oxygen treatmentreceived after her cardiac arrest. The heart showed

e * f *~~~4

K+1- 4 ;_..1<

f ws w s44

Fig. 4 Bone marrow infarction. Amorphous, eosinophiliccell outlines and nuclear debris fill the medullary cavities.The thin walled blood vessels in this field are dilated andcongested but not occluded. Bar = 100 ,um.

Fig. 5 Dilated, congested but not occluded small and largecerebral vessels filed with sickle cells. This was a commonvascular finding in all organs examined. Bar = 100 ,um.

661

on March 23, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.6.659 on 1 June 1985. Dow

nloaded from

Athanasou, Hatton, McGee, Weatherall

A' -a.

'4

4 4

.4 *

* a.

'*1,.

' j

Fig. 6 Anterior pituitary necrosis.Ghost like outlines ofcells ininfarcted area (right) contrast withnormal nucleated pituitary cells(left). Bar = 100 pn.

oedema and myocardial hypertrophy: there were afew haemosiderin granules within myocardial fibres.No infarcts were noted. A single small arteryappeared occluded, but other small and large vesselsexamined were patent, although dilated and filledwith red cells.The spleen was of the small siderofibrotic type

with fibrosis of the splenic pulp, which also con-tained abundant dark brown pigment, mostly ironwith a small amount of calcium. The bone marrowappeared hyperplastic with a predominance of redcell precursors. Recent marrow infarction associatedwith dilated, engorged but not occluded small ves-sels were noted in all the bones examined (Fig. 4).Old bony infarcts were also evident. The livershowed centrilobular necrosis, sinusoidal conges-tion, and erythrophagocytosis by Kupffer cells.There was also considerable haemosiderin deposi-tion in both hepatocytes and Kupffer cells. Noincreased fibrosis or cirrhosis was seen.The kidney showed hyperaemic blood vessels and

enlarged hypercellular glomeruli with engorged,dilated capillary tufts. Afferent arterioles and inter-lobular arteries were also dilated and intensely con-gested. There were scattered, isolated sclerosedglomeruli, and individual enlarged glomerulishowed capsular thickening.The proximal and distaltubules were dilated and contained an eosinophilic,proteinaceous or amorphous granular exudate. Thetubular lining cells also contained haemosideringranules.The brain showed cerebral oedema. No infarction

or thrombosis was seen, but there was considerablecongestion and dilatation of meningeal and cerebralvessels, both large and small (Fig. 5). The anteriorpituitary showed large areas of scattered recent nec-

rosis (Fig. 6). The remaining organs containeddilated vessels filled with red cells but were other-wise histologically unremarkable.

Discussion

Vaso-occlusive crises are a common and potentiallyfatal complication of sickle cell disease. They aremost common in homozygous sickle cell anaemiaand may be recurrent and focal-lfor example,hand-foot syndrome or focal necrosis of individualorgans-or, less often, systemic with generalisedsickling leading to unexpected death.' In both casesthe pathophysiology is thought to be similar. Theincreased viscosity inherent in deoxygenateddeformed sickle cells and oxygenated, irreversiblysickled cells is the major factor responsible for ob-struction of blood flow in the microvasculature.' Inaddition, the elongated rigid sickle erythrocyte isless deformable and flows through capillaries withdifficulty. A vicious cycle is set up whereby theabove factors promote stasis which, in turn, aug-ments local hypoxia and acidosis and leads toincreased sickling and ultimately to microvascularocclusion and ischaemic infarction.9The most striking pathological change in the

above case was the presence of multiple scatteredrecent pulmonary infarcts in both lungs. Microscopicvascular occlusion was not seen and only a handfulof infarcted areas were associated with macroscopicthrombi. It is difficult to reconcile this finding to theabove pathophysiological scheme; in the lungs itwould be expected that gross sickling of unsaturatedblood in the pulmonary arteries would precipitate insitu vaso-occlusion followed by the local addition ofthrombosis after stasis has developed. Moreover,

662

on March 23, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.6.659 on 1 June 1985. Dow

nloaded from

Sickle cell crisis and the chest syndrome

given the history of two similar episodes, the lack offibrous scarring and the absence of organised vascu-lar occlusions in either small or large vessels wassurprising. It would thus appear that extensiveischaemic necrosis can occur in sickle cell anaemiawithout morphological evidence of thrombosis.

In this case capillary engorgement and dilatationwere prominent findings not only in the lungs but inall tissues examined. Capillary engorgement hasoften been described in necropsy reports and pro-tocols of sickle cell anaemia'0-'9 and is thought torepresent capillary stasis. Diggs' proposed thataggregations of entangled sickle cells intermingledwith platelets, white cells, cellular debris, fibrin, andhaemoglobin crystals escape from these dilatedcapillaries, sinusoids, and venules in various parts ofthe body and finally lodge in the pulmonaryarteries.' Such emboli could produce the multipleinfarcts noted, but morphological evidence ofemboli in many of the infarcted areas is lacking.

Kimmelstiel'5 described a case of multiple largeinfarcts of the kidneys, liver, gall bladder, and thebrain without detectable thrombosis in a child withsickle cell anaemia. He also noted capillaryengorgement and was stimulated to review the pub-lished work; he found surprisingly few conclusivereports of vascular thrombosis in sickle cell disease.He suggested that ischaemia in sickle cell crisis couldbe initiated by vascular spasm; this may lead directlyto ischaemic necrosis or, more commonly, be fol-lowed by small vessel dilatation with stasis. Thecapillaries in this phase would appear engorged anddilated. Finally, local ischaemia of tissue and capil-lary walls may result in organic damage to the vesselwalls, degenerative changes, and thrombosis result-ing in superimposed and larger infarction. Diggs'noted that the transient nature of symptoms andsigns in sickle cell crisis would support a vasospasticprocess.

Kimmelstiel's hypothesis could neatly explain thelung findings in our case. Areas of pulmonary infarc-tion unassociated with thrombosis are presumablydue to vasospasm, and areas of normal lung contain-ing engorged small vessels represent areas whichhave remained viable after the vasospastic episode.Where thromboses are seen, these form on the basisof an accompanying increase in blood viscosity andthe rheological inefficiency of the sickled cells; localischaemia of tissue and capillary walls resulting inorganic damage to the vessel wall with degenerativechanges would also favour thrombus formation.Acute pulmonary disease is a common and serious

consequence of sickle cell disease.26 A common pre-sentation is that of an acute febrile episode associ-ated with the development of a pulmonary infiltrateon chest radiograph. Patients with sickle cell

663

anaemia are at greatly increased risk of infectionwith certain bacteria, particularly pneumococcus;but in the acute chest syndrome often no organismis isolated.6 20-23 Margolies20 suggested that acutepulmonary changes in sickle cell anaemia are due tomultiple infarcts producing a pneumonia like pic-ture. Oppenheimer and Esterley7 found throm-boemboli in most patients with sickle cell diseaseand noted that pneumonia did not appear to begreater than that in age matched contols. Severalrecent combined microbiological and clinical studiesfavour pulmonary intravascular sickling or pulmo-nary infarction as the cause of the chest syn-drome.21-23Our case showed typical features of the acute

chest syndrome.23 There was a history of cough andchest pain followed by rapidly developing featuresof bilateral consolidation. Sputum and blood cul-tures gave consistently negative results; pneumococ-cal antigens were not detected; and there was noevidence of chest infection histologically. The clini-cal and radiological features in this case would thusappear to be accounted for by engorgement of thepulmonary vasculature with sickle cells and pulmo-nary infarction rather than infection. As discussedearlier, the striking absence of thrombosis toaccount for the large numerous areas of pulmonaryinfarction would appear to suggest that intravascularsickling alone, possibly preceded by an episode ofvasospasm, can lead to pulmonary infarction. Estab-lished thrombosis is not necessary for pulmonaryinfarction to occur but may result because of theincrease in blood viscosity and the rheological ineffi-ciency of sickled cells, which favour vascular stasis.There are a number of factors which make the

lung a favourable site for sickling. The mixed venousoxygen tension in the pulmonary capillary bed isabout 40 mm Hg, normally supporting an oxygensaturation of 70%. Even this physiological degree ofvenous saturation is compatible with widespreadsickling in patients with homozygous sickle cellanaemia, as shown by the fact that about 30-60%of in vivo sickling occurs on the venous side com-pared with only 20% on the arterial side.24Moreover, the lower pH of venous blood would alsofavour sickling.25 In adults or adolescents withhomozygous sickle cell anaemia, gas exchange isoften abnormal in clinically stable periods, and thealveolar-arterial pO2 difference is widened mainlybecause of excessive physiological shunting of bloodand an abnormal degree of heterogeneity of ventila-tion to perfusion ratios in the lung.26 Sickle lungsyndrome has occasionally been associated withmild upper respiratory tract infections.3 This cancause bronchiolar narrowing, which would lead tosmall zones of hypoventilation and a consequent

on March 23, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.6.659 on 1 June 1985. Dow

nloaded from

664

increase in the physiological shunt. The effect ofsuch anoxic alveolar areas would be to preserve thesickle form of the erythrocyte in the pulmonary ven-ous circulation and further decrease the velocity offlow in poorly ventilated alveolar areas.27

Massive pituitary necrosis, as occurred in our

case, has not previously been described anatomicallyin sickle cell crisis. It is associated with obstetric andnon-obstetric shock and has been reported inpatients who were maintained on mechanical ven-tilators before they died.28 Our patient had a cardiacarrest and was maintained on mechanical ventilatorsfor two days before her death, and so care must betaken in attaching too much importance to thisfinding. It is interesting to note, however, that thereis one clinically confirmed case of Sheehan's syn-drome developing in a patient with sickle cellhaemoglobin C disease29 and that postpartum pitu-itary necrosis is thought to arise from severe vaso-

spasm.28

We are grateful to Dr MS Dunnill for helpful discus-sions and to Miss L Watts for typing the manuscript.

References

Diggs LW. Sickle cell crisis. Am J Clin Pathol 1965;44: 1-19.2 Davies SC, Hewitt PE. Sickle cell disease. Br J Hosp Med

1984;31:440-4.Anionwu E, Walford D, Brozovic M, Kirkwood B. Sickle cell

disease in a British urban community. Br Med J1981;282:283-6.

4Rickles FR, O'Leary DS. Role of coagulation system inpathophysiology of sickle cell disease. Arch Intern Med1974; 133:635-41.

5Barrett-Connor E. Bacterial infection and sickle cell anemia.Medicine 1971;50:97-112.

6 Barrett-Connor E. Acute pulmonary disease and sickle cellanemia. Am Rev Resp Dis 1971; 104: 159-65.

7Oppenheimer EH, Esterley JR. Pulmonary changes in sickle celldisease. Am Rev Respir Dis 1970; 103:858-9.

Chien S, Usami S, Bertles JF. Abnormal rheology of oxygenatedblood in sickle cell anemia. J Clin Invest 1970;49:623-34.

9 Klug PP, Lessin LS, Radice P. Rheological aspects of sickle cell

Athanasou, Hauton, McGee, Weatherall

disease. Arch Intern Med 1974; 133:577-90.'0 Sydenstricker VP, Mulherin WA, Houseal RW. Sickle cell

anemia: report of two cases in children with necropsy in onecase. Am J Dis Child 1923;26: 132-54.

"Steinberg B. Sickle cell anemia. Arch Pathol 1929;9:876-97.2 Diggs LW, Ching RE. The pathology of sickle cell anemia. South

Med J 1934;27:839-45.Hughes JG, Diggs LW, Gillespie CE. The involvement of the

nervous system in sickle cell anemia. J Pediatr 1940;17:166-84.

14 Tomlinson WJ. Abdominal crises in uncomplicated sickle cellanemia. Am J Med Sci 1943;202:722-41.

"Kimmelstiel P. Vascular occlusion and ischemic infarction insickle cell disease. Am J Med Sci 1948;216: 11-19.

16 Edington GM. The pathology of sickle cell disease in WestAfrica. Trans R Soc Trop Med Hyg 1955;49:253-67.

Edington GM. The pathology of sickle cell haemoglobin C dis-ease and sickle cell anaemia. J Clin Pathol 1957; 10: 182-6.

Thorburn MJ. The pathology of sickle cell anaemia in Jamaicanadults over 30. Trans R Soc Trop Med Hyg 1969;69:98-111.

'9 Diggs LW. Anatomic lesions in sickle cell disease. In: AbramsonIT, Bertles JF, Wethers DL, eds. Sickle cell disease. Diagnosis,management, education and research. St Louis: CV Mosby Co,1973:230.

Margolies MP. Sickle cell anaemia. Medicine 1951;30:357-443.21 Charache S, Scott JC, Charache P. Acute chest syndrome in

adults with sickle cell anaemia. Arch Intern Med1979; 139: 67-9.

22 Young RC, Castro 0, Baxter RP, et al. The lung in sickle celldisease: a clinical overview of common vascular, infectious andother problems. J Natl Med Assoc 1981;73: 19-26.

21 Davies SC, Win AA, Luce P, Riordan JF. Acute chest syndromein sickle cell disease. Lancet 1984;i:36-8.

24 Jensen WN, Rucknager DL, Taylor WJ. In vivo study of thesickle cell phenomenon. J Lab Clin Med 1960;56:854-65.

25 Finch CA. Pathophysiological aspects of sickle cell anaemia. AmJ Med 1972;53: 1-6.

26 Bromberg PA. Pulmonary aspects of sickle cell disease. ArchIntern Med 1974; 133:652-8.

27 Moser KM, Shea JG. The relationship between pulmonaryinfarction, cor pulmonale and the sickle states. Am J Med1957;22:561-80.

26 Ezrin C, Kovacs K, Horvath E. Pathology of theadenohypophysis. In: Bloodworth JMB Jr, ed. Endocrinepathology. Baltimore: Williams and Wilkins, 1982:112.

29 Adadevoh BK. Haemoglobin sickle cell disease and Sheehan'ssyndrome. Br J Clin Prac 1968;22:442-3.

Requests for reprints to: Dr N Athanasou, University ofOxford, Nuffield Department of Pathology, John RadcliffeHospital, Oxford OX3 9DU, England.

on March 23, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.6.659 on 1 June 1985. Dow

nloaded from