Pediatr Radiol (1999) 29: 646±661 Ó Springer-Verlag 1999

John J. Crowley Sharada Sarnaik

Received: 28 January 1999 Accepted: 23 April 1999

)

J. J. Crowley ( ) Department of Pediatric Imaging, Children's Hospital of Michigan, 3901 Beaubien Boulevard, Detroit, MI 48201-2196, USA S. Sarnaik Sickle Cell Center, Children's Hospital of Michigan, 3901 Beaubien Boulevard, Detroit, MI 48201-2196, USA

Imaging of sickle cell disease

Abstract Sickle cell disease is an important health care issue in the United States and in certain areas in Africa, the Middle East and India. Although a great deal of progress has been made in understanding the disease at the molecular and pathophysiologic level, specific treatment which is safe and accessible for most patients is still elusive. Going into the next millennium, the management of this disease is still largely dependent on early diagnosis and the treatment of complications with

Introduction Considering the clinical importance of sickle cell disease [1], it is surprising to learn that the first description of the disorder was not published until Herrick's report of a 20-year-old dental student in 1910 [2]. In 1949, the abnormal electrophoretic mobility of sickle hemoglobin (2, bs2) was noted, prompting Linus Pauling and his colleagues to christen sickle cell anemia ªa molecular diseaseº [3]. By 1961, the biochemical defect, the substitution of valine for glutamic acid at the sixth position of the b-globin chain, had been defined [1]. In its deoxygenated state, sickle hemoglobin polymerizes and distorts red-blood-cell shape, causing oxidative damage, cellular dehydration, abnormal phospholipid symmetry, and increased adherence to endothelial cells [4, 5]. These changes in flow properties cause vasoocclusion, which is episodic and recurrent and leads to chronic organ damage. Red-cell survival is decreased and a chronic hemolytic anemia results. It is sobering to reflect that in the United States alone there were 75,000 hospitalizations per year for patients with sickle cell disease between the years 1989 and

supportive care. Thus, diagnosis and evaluation of the complications of the disease are crucial in directing clinical care at the bedside. Modern imaging modalities have greatly improved, and their application in the patient with the sickling disorders has enhanced the decision ± making process. The purpose of this article is to review the clinical aspects of common complications of the disease and to discuss imaging approaches which are useful in their evaluation.

1993 [6]. This yields an estimated cost of $ 470,000,000 per year (1996 dollars); in 36 % of cases, government programs were listed as the expected principal source of payment [6]. Although clinical manifestations of sickle cell disease are legion, in this review we will concentrate on the more common and clinically significant complications that the pediatric radiologist is likely to encounter. In the United States, one large study showed a prevalence of the sickle cell trait of 8.6 % of those of African descent [7]. The genetic mutations for sickle cell disease are believed to have evolved in Africa with heterozygotes having some resistance to malaria [8]. Sickle cell trait occurs at a prevalence of 10±30 % in central Africa [8]. The disease is also found throughout the Middle East and India; 20 % of the population of some parts of Saudi Arabia have sickle cell trait [9]. Sickle cell disease refers to those patients who are homozygous for the sickle cell gene. Over 95 % of their hemoglobin will be sickled hemoglobin, with the remainder made up of some trace hemoglobins. In those patients with sickle cell trait 40±45 % of their hemoglobin will be sickled [10]. Sickle cell trait is fortunately

647

usually a clinically benign condition [11]. There are occasional reports of sudden death during extreme exertion, such as in military recruits [12]. Renal concentration ability may also be impaired with polyuria and nocturia, and splenic infarction at high altitudes has also been reported [11]. Fortunately, serious complications are rare. Other minor hemoglobinopathies include Hb SC disease and Hb S/beta thalassemia. Generally, these minor hemoglobinopathies pursue a similar but milder course to homozygous sickle cell disease. In general, they have no characteristic radiological or clinical presentations apart from their splenic manifestations, which we will discuss later.

Thoracic complications Clinical significance The acute chest syndrome (ACS) is a complication of sickle cell disease characterized by pleuritic chest pain, fever, rales on lung auscultation, and pulmonary infiltrates on chest ray [13]. A commonly accepted definition of the syndrome is the occurrence of a new infiltrate in a chest X-ray with chest pain and/or pulmonary symptoms [14]. The ACS may be due to infection, infarction, fat embolism, or a combination of all three [13]. It affects nearly one-third of all patients with sickle cell disease and is a frequent cause of death [13]. The incidence is strongly but inversely related to age, being highest in children 2± 4 years of age and decreasing to its lowest level in adults [13]. Factors shown to increase the rate of ACS are a high, steady-state Hb level and a low fetal Hb level. Although a high rate of ACS is associated with the worse prognosis, it does not follow that the patients will die of this complication; more likely it is a marker of disease severity [13]. Lung involvement constitutes a paramount determinant of survival with repeated episodes of ACS predisposing to chronic lung disease and early death [15, 16]. Acute pulmonary disease has become the most common cause of death and the second most common cause of hospitalization in sickle cell disease [17]. Recently, a large cooperative prospective study reported their experience with 3,751 patients [16]. Patients were enrolled from birth to 66 years of age. There were 1,722 episodes in 939 patients, of whom 738 were Hb SS, 127 were Hb SC, and 74 were Hb S beta thalassemia. ACS was most common in winter, with children having the most striking increase [16]. Children were usually febrile and had a cough but rarely presented with chest pain [16]. In contrast, adults were often afebrile but had severe pain. Upper lobe disease was more common in children compared to multilobe disease in adults [16]. Bacteremia was demonstrated in only 3.5 % of patients but was found in 14 % of infants

versus 1.8 % of patients over 10 years of age [16]. Overall, in children ACS was milder and more likely due to infection, whereas in adults ACS was severe, associated with pain, and had a higher mortality [16]. The etiology of the ACS is unclear [1, 12, 15, 16, 18]. Martin and Buonomo [19] found no etiology in 87 % of episodes and an infectious agent is rarely identified even with aggressive bronchoscopy and cultures [1, 19±21]. It has long been speculated that in situ thrombosis is the etiology of ACS [13, 22]. Data from the National Acute Chest Study confirmed rapid improvement in oxygenation in several hundred patients transfused for deterioration of pulmonary disease, although how transfusion produces such immediate benefit is unknown [17, 23]. Bhalla et al. [24] reported detecting microvascular occlusion with high-resolution CT in a small series of patients. However, most episodes resolve without anticoagulants or antifibrinolytics, which has been seen as arguing against the role of thrombosis in these cases [1]. Interest has also centered on a fat embolism in ACS, and bronchoalveolar lavage has detected fat-laden macrophages, suggesting fat embolism in a significant proportion of cases [25]. Recently, interest in the role of bone-marrow fat emboli has been spurred by the detection of serum markers for fatty emboli [26]. These have been shown to be dramatically elevated in patients with ACS, but not in patients with vasoocclusive crises or non-sickle cell disease patients with pneumonia, and the levels have also been shown to correlate with the clinical severity of the syndrome [26]. However, in most cases, the syndrome has multifactorial etiology and is managed by antibiotics, incentive spirometry, and judicious transfusion and respiratory support. Imaging approaches As previously described, any new chest infiltrate in a patient with sickle cell disease is, by definition, ACS [14]. In a recent series, in 46 % of episodes, the initial chest X-ray was normal [19]. The same series noted two separate radiographic patterns. In most cases (87 %), no etiology was ever found and these patients had lowerlobe airspace disease with rapid onset and rapid resolution. The median age of this group was 14 years. In the second group (13 %), an identifiable etiology was found that was usually bacterial pneumonia; these patients had a prolonged radiographic course with improvement not seen until 4 days post-therapy. The median age in this group was 2 years [19]. Chest radiographs are, however, of limited value in patient management, as the radiographic findings may lag behind the clinical symptoms and may change with bewildering speed (Fig. 1). The CT appearance of the lungs in patients with ACS was reported in a small series of 10 patients by Bhalla et al. [24]. In 9 of 10 patients they noted a paucity or ab-

648

a

b

Fig. 1 a, b A 6-year-old boy admitted with fever and dyspnea. a Initial radiograph showed only cardiomegaly and mild vascular conjestion. b The following day there is complete consolidation of the left lung

sence of small vessels in the periphery of the lung, which they attributed to microvascular occlusion. They also found a ground-glass attenuation, presumed to be due to hemorrhagic edema [24]. However, these findings are of little practical significance, as bacterial pneumonia can coexist with microvascular occlusion and all patients will be treated with IV analgesics, antibiotics, and judicious hydration regardless of the CT findings. Chest radiography has been reported to underestimate the degree of vascular damage, as the depth of hypoxia is generally out of proportion to the degree of consolidation noted on the chest radiograph [27]. Pulmonary scintigraphy can reveal the status of the pulmonary circulation and can do so without the risk of pulmonary angiography, as contrast material may aggravate the sickling process [24, 27]. Nuclear medicine scintigraphy has shown defects in both large and small pulmonary vessels and when the large vessels are involved, the scintigraphic appearance mimics pulmonary embolism [27]. In this case, a source of emboli such as pelvic or leg deep-venous thrombosis should be sought before the perfusion defects are ascribed to sickle cell disease, as the two conditions may coexist [28]. In sickle cell disease, the perfusion defects often resolve rapidly with only supportive therapy and without anticoagulants [27]. Even the demonstration of perfusion defects has limited prognostic significance, as such findings may be primary because of sickling of erythrocytes [27] or secondary to a range of disorders such as pneumonia or fat embolism related to infarction of necrotic bone [29]. It is uncertain how frequently the latter condition occurs, partly because it is difficult to diagnose. However, it is believed to occur more frequently in adults and to predis-

pose to the pulmonary hypertension seen in this condition [29]. Thus the demonstration of perfusion defects on scintigraphy does not establish sickling of erythrocytes within the circulation as the primary pathologic event in any patient. In recent years there has been interest in rib infarcts as a cause of the acute chest pain in some patients with ACS [30, 31]. It is proposed that rib infarcts give rise to local soft-tissue swelling and pleuritic pain [30]. Splinting and hypoventilation secondary to pain then lead to atelectasis and the radiographic changes of pneumonia, with effusion following on from pleural or pulmonary involvement [30]. In summary, the ACS in sickle cell disease as currently defined may represent varied pathophysiology; infectious pneumonia (bacterial, viral, or parasitic), embolic or vasoocclusive disease, or atelectasis from reactive airway disease are all diagnostic considerations. Splinting from rib infarcts causing stasis may also be a contributing factor. Although CT scanning and nuclear medicine scintigraphy have been used to evaluate occlusive or thrombotic disease, their role in the clinical management of the patient is limited. It does not appear critical to make this distinction because anticoagulants are not useful in this situation. Bronchoscopy and ultrasoundguided needle aspiration of pleural fluid, when feasible, may be of use to direct appropriate antimicrobial therapy. In some cases, bronchoscopy is therapeutic, particularly in the aspiration of inspissated mucous plugs. Imaging beyond plain radiographs is rarely necessary, and even these are of limited use as they often correlate poorly with the patient's clinical condition.

Splenic complications Clinical significance The spleen is uniquely susceptible to damage in Hb SS [1] because of its slow, tortuous microcirculation. In the first 6 months of life, the red pulp of the spleen becomes

649

Fig. 2 a, b A 6-year-old girl with acute abdominal pain and hypotension. Note splenic shadow extends into the pelvis and displaces the stomach medially

a congested with sickled erythrocytes that may lead to functional asplenia or acute splenic sequestration [1]. Gradually small hemorrhages and infarctions develop, the spleen becomes smaller, and is eventually replaced by fibrous tissue, a process termed autosplenectomy [1]. This splenic dysfunction is associated with lethal bacterial infection that is a common cause of death in sickle cell disease. Splenomegaly is rare in adults with sickle cell anemia, but is common in other sickle hemoglobinopathies, such as Hb SC and Hb S beta thalassemia [32, 33]. Splenomegaly must be present for a splenic sequestration crisis to develop and, therefore, this complication is usually seen before the age of 6 years in sickle cell anemia, but may occur in older children or adults in the minor sickle hemoglobinopathies [32±34]. Splenic hypofunction is present in nearly 30 % of infants with sickle cell anemia at the first birthday and in 90 % by age 6 years, accounting for the high risk of encapsulated polysaccharide bacterial septicemia [35]. Splenic function is preserved in most cases of Hb SC disease and Hb S beta thalassemia [35]. Splenic sequestration occurs when there is a sudden accumulation of blood within the spleen resulting in severe anemia and hypovolemia [36, 37]. These events occur in young children usually under the age of 6 years who have not as yet developed splenic fibrosis. Criteria for acute splenic sequestration include a decrease in hemoglobin of at least 2 g/dl, an acutely enlarging spleen, evidence of bone-marrow compensation such as reticulocytosis, or an increased number of nucleated red cells on blood films [35]. The hematocrit may drop rapidly leading to hypovolemic shock [36]. Treatment is supportive care, focused on correction of anemia and hypovolemic shock by red cell transfusions [36]. Cardiovascular collapse and death will occur in

b 10±30 % of children, most commonly between 6 months and 3 years of age and frequently precipitated by a viral infection [37, 38] (Fig. 2). Up to 50 % of children with one episode will have a second episode, usually within 2 years [1, 39]. Following acute splenic sequestration, the spleen may remain enlarged with clinical evidence of hypersplenism and this is an indication for splenectomy [1, 40]. Indications for splenectomy in sickle cell disease are controversial. Patients with a first episode of splenic sequestration have been reported to have a high recurrence rate, with additive morbidity and mortality. Some clinicians advocate a splenectomy after a first episode since the organ is expected to be nonfunctional (from auto splenectomy) in a vast majority of patients. Others adopt a more conservative approach and institute a temporary transfusion regimen. This preserves residual splenic function (if present) and avoids the risks of abdominal surgery. Acute splenic sequestration rarely occurs in the first 5 months of life, although fatal cases have been reported as young as 2 months of age [37]. Unfortunately, acute splenic sequestration crisis frequently goes undetected until the child is in profound shock [37], and the first clue may be the detection of an enlarged spleen either on clinical examination or on a plain abdominal X-ray (Fig. 2) [41]. In later childhood, splenomegaly is less common and splenic sequestration becomes rare [42]. There are, however, rare case reports of acute splenic sequestration in adults with homozygous sickle cell disease [43]. In older children, the spleen is small and fibrotic in patients with sickle cell anemia. This contrasts with hemoglobin SC disease and sickle beta thalassemia where more than 50 % of individuals have an enlarged spleen, although splenic complications, such as acute splenic se-

650

Skeletal complications Clinical significance

questration, are relatively uncommon in this disease in contrast to homozygous sickle cell disease [42]. Recently, there has been interest in partial splenectomy with initially promising results [44].

Bone pain is a common symptom in sickle cell disease, and the differentiation of infarction from osteomyelitis is a difficult clinical problem. It must be borne in mind that infarction is more common than infection; one report describes a 50-fold difference [47] (Figs. 4±7). When osteomyelitis does occur, it is frequently caused by Salmonella (Fig. 8). A recent review of the literature concluded that Salmonella was the most common cause of osteomyelitis in patients with sickle cell disease both in developing countries and in the developed world and that its relative incidence was more than twice that of Staphylococcus aureus [48]. The management of uncomplicated bone infarction differs from that of osteomyelitis. The former requires supportive care, such as prevention of dehydration (usually with intravenous fluid) and pain medication. The latter requires an additional commitment of weeks of intravenous and oral antibiotics. Thus there is a need for differentiating osteomyelitis from simple infarction.

Imaging approaches

Imaging approaches

Walker and Serjeant [45] have described focal echogenic lesions of the spleen on ultrasound (US) in patients with sickle cell disease that were unrelated to symptoms and were unchanged at a 12-month followup. These are probably benign and do not warrant further investigation [45]. Levin et al. [46] described six patients (five with Hb SS disease and one with sickle beta thalassemia) in whom rounded splenic masses represented preserved, functioning splenic tissue. On US, these were hypoechoic relative to the rest of the spleen, showed low attenuation on CT, and had imaging characteristics of normal spleen on MRI. The splenic masses demonstrated uptake of 99mTc-sulfur colloid on liver spleen scans but showed no uptake of 99mTc-MDP on bone scans. 99m Tc-sulfur colloid spleen scans are often performed to evaluate residual reticuloendothelial splenic function (Fig. 3). It has been demonstrated that splenic dysfunction may be reversible with transfusions. Scans may be useful in the clinical decision regarding splenectomy. Patients with hypersplenism without reticuloendothelial 99m Tc-sulfur colloid uptake in the spleen are candidates for splenectomy; on the other hand, if residual function is seen, a decision may be made to postpone the procedure.

It has long been known that either a hot or cold spot on a nuclear medicine bone scan may represent either infection or infarction [49]. One approach to differentiate the two is to combine bone marrow imaging with technetium 99m sulfur colloid with a 99mTc-diphosphonate bone scan [50]. Decreased uptake on bone-marrow scans in a patient with sickle cell disease and bone pain strongly suggests infarction, whereas normal uptake strongly suggests osteomyelitis [50]. Unfortunately, the photondeficient area within bone may be seen early in the course of osteomyelitis as well as in bone infarction, and the increased uptake seen in the healing infarct may also be seen in osteomyelitis [51]. A sterile and an infected infarction may also be indistinguishable on bone imaging [51]. A combination of technetium and gallium scintigraphy has also been used in an attempt to differentiate infarction from infection in sickle cell patients [52]. Mandell [49] in an excellent discussion on this subject points out that either a hot or cold spot on a nuclear medicine bone scan may be because of either osteomyelitis or avascular necrosis. One approach is to combine bonemarrow scanning (99mTc-sulfur colloid) and 99mTc-MDP bone scan. An acute infarct will always show decreased marrow uptake, whereas a normal marrow scan in an area of suspected acute osteomyelitis on the bone scan is likely because of infection [49]. Gallium scanning may be helpful, as normal or only slightly increased gallium uptake is strongly against osteomyelitis [35]. If a 99mcTcMDP bone scan is cold in acute infection, gallium may

Fig. 3 A 4-year-old boy with history of prior splenic sequestration. Technetium-labeled sulfur-colloid scan shows good function in the persistently enlarged spleen

651

Fig. 4 a, b A 12-year-old boy with multiple episodes of bone pain. Frontal and lateral radiographs of the left femur show diffuse patchy sclerosis throughout the shaft because of healed bone infarcts Fig. 5 A 6-year-old girl with right hip pain. There is a subchondral lucency in the right femoral head indicating early avascular necrosis Fig. 6 A 7-year-old girl with dyspnea. The heart is enlarged with engorgement of central pulmonary vessels. There is avascular necrosis of each humeral head Fig. 7 Typical ªcodfishº vertebrae in a 15-year-old boy caused by microinfarctions of the vertebral end plates Fig. 8 a±c A 10-year old girl with fever, elevated white count, and left leg pain. a, b Frontal and lateral radiographs of the left tibia show mottled lucencies throughout the shaft. A surgical biopsy site is evident, which yielded Salmonella (arrows). c Same patient 18 months later. Dense sclerotic change in the shaft represents treated osteomyelitis

4a

4b

5

7

6

8a

8b

8c

652

Fig. 9 a, b An 11-year-old boy showing the classic ªhair-onendº appearance of hematopoiesis within the medullary cavity of the skull

a

Fig. 10 A 12-year-old girl. Longitudinal image of the gallbladder shows multiple, echogenic, shadowing foci representing numerous gallstones in the dependent portion of the gallbladder

show increased uptake in osteomyelitis and normal uptake in infarction [49]. When the distribution of activity is similar, higher gallium uptake compared to bone-tracer uptake implies infection. The differentiation of osteomyelitis from infection has also been attempted by using tagged white cells with initially promising results [53]. Stark et al. [54] described identifying subperiosteal abscesses in two patients with bone pain and fever and equivocal plain radiographic and bone-scan findings using contrast-enhanced CT. The authors did not study any patients with bone infarcts and were therefore unable to state whether subperiosteal fluid collections could occur in the absence of infection. CT scanning was, however, helpful to guide aspiration.

b There has also been interest in differentiating osteomyelitis from infarction by using MRI. One study confined to children found that gadolinium-enhanced MRI was totally unable to differentiate acute osteomyelitis from acute bone infarct [55]. Soft-tissue changes, pattern of enhancement, and periosteal reaction were the same in each disease. MRI was, however, helpful in determining the anatomic site and extent and guiding aspiration [55]. Recently, enthusiasm for nuclear medicine and magnetic resonance scans has waned, as they both have been found to be fraught with both false-positive and false-negative results [47, 56]. In the rare case of osteomyelitis, the patient is usually highly febrile, toxic, and has a left shift in the leukocyte count [47]. In doubtful cases, aspiration may be indicated. However, unless an infection is proven by culture or strongly suspected on clinical grounds, a prolonged course of antibiotics is not indicated [47]. In summary, imaging differentiation between osteomyelitis and simple infarction is not reliable. Laboratory results such as white count and sedimentation rate are also unreliable. The only definitive diagnostic test is a positive culture from blood or bone puncture. Failing this, most clinicians will use clinical judgment in making this distinction, bearing in mind the relative rarity of osteomyelitis compared to infection. The current major role of bone imaging is to guide placement of the needle so as to obtain a reliable specimen for culture. In addition to the problem of differentiating osteomyelitis and infection, increased hematopoiesis caused by profound anemia may produce bizarre radiographic patterns (Fig. 9).

653

Fig. 11 a, b Same patient as Fig. 10. One year later the patient presented with severe upper abdominal pain. Frontal image from HIDA scan obtained at 45 min (a) and lateral image obtained at 60 min (b) after radiotracer injection show nonvisualization of the gallbladder. Acute calculous cholecystitis confirmed at surgery

11 a

Fig. 12 a, b A 10-year-old girl with acute right upper abdominal pain during a sickle crisis. Ultrasound showed gallbladder wall thickening of more than 1 cm. Acute acalculous cholecystitis confirmed at surgery

11 b

12 a

Liver and biliary system Gallstones are present in 10±55 % of American children with sickle cell disease, depending on age [57] (Fig. 10). The incidence shows a wide variation in different geographical regions of the world [58]. The differentiation of abdominal pain crisis from cholecystitis may be difficult (Figs. 11, 12); in some centers patients undergo screening sonography and prophylactic cholecystectomy if gallstones are found [57]. There are five specific hepatobiliary syndromes affecting patients with sickle cell disease [59]. These are viral hepatitis, hepatic crisis, cirrhosis, cholelithiasis with cholecystitis, and intrahepatic cholestasis. This latter syndrome is characterized by sudden onset of severe right upper-quadrant pain, progressive hepatomegaly, coagulopathy with hemorrhage, and extreme hyperbilirubinemia [59]. Most patients succumb [60, 61]. Treatment is supportive therapy with emphasis on correcting the coagulopathy and blood-plasma transfusion [60, 61]. The increased incidence of both gallbladder sludge and gallstones in patients with sickle cell disease was

12 b well documented soon after US came into widespread clinical use [62, 63]. The fear is that gallstones may cause cystic duct obstruction and acute cholecystitis; however, acute cholecystis may occur in the absence of stones in critically ill children with the ACS or acute pain crisis (Fig. 12). The increased incidence of predominately pigmented gallstones is believed to be caused by accelerated hemolysis; biliary sludge is believed to result from supersaturation of gallbladder bile-forming crystals of cholesterol or bilirubinate [64]. It is further presumed that crystalline particles in sludge aggregate to form discrete stones [64]. However, stones may form without prior sludge [64], and the precise relationship of biliary sludge to biliary stones and biliary stones to acute cholecystitis, biliary obstruction, and abdominal pain has long been controversial. Unfortunately, recent studies have only deepened the confusion. In a retrospective review of 75 children with sickle cell disease, Winter et al. [65] identified 17 patients with biliary sludge, 9 of whom also had stones and had elective cholecystectomy. In eight patients with sludge alone, stones developed in every patient within 30 months, and the authors recom-

654

mended cholecystectomy in any patient with sickle cell disease and sludge [65]. In contrast, a prospective study of 429 Jamaican children followed with annual US examinations from 1987 to 1995 identified gallbladder sludge in 17 patients [64]. Another 12 patients went on to develop stones, sludge disappeared in 4 patients, and in 1 patient, sludge persisted. No patient with sludge had symptoms referable to the biliary system [64]. The same study detected 98 patients with stones, yet only 4 developed symptoms requiring cholecystectomy, prompting the authors to recommend a conservative policy both to gallbladder sludge and asymptomatic gallstones [64]. Although the patients in the study of Winter et al. had some abdominal symptoms whereas the study of Walker et al. examined asymptomatic patients, an editorial comment in The Journal of Pediatrics conceded that the natural history of biliary sludge and biliary stones in patients with sickle cell disease remains controversial [57].

CNS disease Clinical approach The peak incidence of death in patients with sickle cell disease occurs between 1 and 3 years of age with the primary cause of death being bacterial infection [66]. The next most frequent cause of death is cerebrovascular accident (CVA), accounting for 12 % of pediatric deaths from sickle cell disease [66]. In a Jamaican cohort study of 310 children with homozygous S followed since birth, 7.8 % (17 of 310) had a CVA by the age of 14 years, 2 having subarachnoid hemorrhage and the remaining 15 presumed to be caused by infarction [67]. Six of the 13 (46 %) patients who survived the initial event had a recurrent CVA [67]. There was a median interval of 9 months from initial event to the recurrence. There were six deaths, two during the acute event and four after recurrence [67]. Of interest was the finding that a sudden decrease in hemoglobin, which may be caused by aplastic crisis, acute splenic sequestration, or pulmonary sequestration, was a risk factor for stroke [67]. A number of studies have shown that the high rate of recurrence after a first CVA can be prevented with a chronic transfusion regimen [68]. Recent interest has been sparked in primary prevention of stroke for highrisk patients who can be identified and selected for this successful preventive therapy before an overt clinical stroke has occurred [69]. Abnormal neuropsychologic testing has been related to imaging finding of cerebrovascular disease before overt neurologic symptoms appear [70].

Imaging approaches Recently Moran et al. [71] published a comprehensive review of imaging of CNS complications of sickle cell disease. The four modalities employed were angiography, MR imaging, CT scanning, and US. Although use of catheter angiography has decreased with the advent of MR imaging, it remains the gold standard for the assessment of intracranial vessels (Figs. 12, 13). The risk of angiography is that of precipitating intravascular sickling. In a compiled series of 174 arteriograms in patients with sickle cell disease, however, no permanent neurological events were recorded [71] (Figs. 13, 14). Patients should be prepared with transfusions designed to decrease Hb S levels to < 30 % prior to the procedure. There are three patterns of vascular disease: (a) proximal branch occlusion or stenosis, (b) distal branch occlusion, and (c) aneurysm [71]. At our institution we are frequently asked to carry out cerebral arteriography in cases where a screening MRA examination is positive, as the finding of large-vessel disease may prompt aggressive and life-long transfusion therapy [68]. Overall, the use of catheter angiography seems to be declining, with more reliance being placed on MR angiography, which we will discuss later. There remain, however, certain indications for catheter angiography, which include cases where it is important to detect small-vessel occlusion or cases where surgical revascularization is anticipated [72]. CT scanning is widely used to detect both parenchymal and subarachnoid blood, but is less reliable for detection of cerebral infarction [71, 73], particularly when it is acute and nonhemorrhagic. CT has been largely supplanted by MRI. The administration of iodinated contrast medium raises concerns about precipitating sickling. MRI has long been known to be more sensitive than CT for detecting acute infarcts [72]. Recently, diffusion and T2-weighted echoplanar MR imaging have been reported to have even greater sensitivity than routine MRI for differentiating acute hemorrhagic from acute nonhemorrhagic stroke [74]. During the 24 h after onset, conventional MRI is less sensitive than CT for detecting hemorrhage [73, 74]. Diffusion-weighted MR imaging has also been shown to be superior to both CT and conventional MR images in the detection of acute infarcts less than 6 h old [71, 74, 75], the ªwindowº within which therapy must occur if cerebral infarction is to be averted. In addition, although CT has been reported as superior to MR imaging in the detection of acute subarachnoid blood, new MR sequences such as fluid-attenuated inversion recovery (FLAIR) have been reported to be excellent for detection of subarachnoid blood and even superior to CT for blood in the posterior fossa [76]. MR angiography is an exciting technique for evaluating blood flow to the brain in a noninvasive manner (Fig. 15). It has achieved some acceptance in the evalua-

655

Fig. 13 a T2-weighted magnetic resonance image of 8-year-old girl with dense right-sided hemiparesis shows extensive ischemic change in the left hemisphere. b, c Frontal and lateral views of a left common carotid injection show complete occlusion of the left internal carotid artery. d Injection of the right internal carotid artery shows extensive filling of the left hemisphere through a patent arterior communicating artery. e Injection of the left vertebral artery shows a widely patent posterior communicating artery with filling of the pericallosal artery

a

b

c

d

e

656

Fig. 14 a A 4-year-old girl with left-sided weakness. Protondensity-weighted magnetic resonance image shows evidence of ischemia in the white matter of each cerebral hemisphere. Injection of left (b) and right (c) internal carotid artery shows an area of narrowing within the carotid siphon more marked on the left (arrows)

b

a

tion of the extracranial carotid artery; however, it is not without problems in the depiction of intracranial vascular abnormalities [77]. Atlas has rightly cautioned that the ªease of performanceº of this test and the ease in interpreting it by physicians ªcombine to make intracranial MR angiography a potentially highly dangerous diagnostic test in some instances, notably in patients suspected of having intracranial aneurysmsº [78]. Nevertheless, MR angiography has an established role in imaging the extracranial carotid-artery bifurcation [77, 78] and is excellent for detecting large-vessel occlusion [71]. A recent study of MRA of the intracranial circulation carried out in 14 patients with homozygous Hb SS disease and 8 patients with Hb SC disease found that the finding of a segment of narrowing, greater than 6 mm long, at the internal carotid artery bifurcation associated with reduced distal flow, was strongly associated with subclinical cerebral infarcts [79]. Estimating the degree of vessel stenosis with MR angiography remains somewhat controversial [71, 78, 80], particularly in the presence of turbulent flow. Technical advances, particularly with post-processing techniques, have increased the sensitivity for detecting intracranial aneurysms to 75 % [77]. Although it is the policy of this institution to carry out MR angiography as a screening test and not to commit a child to surgery or life-long transfusion therapy without a catheter angiogram, many institutions are much less aggressive in the use of catheter angiography. It is our impression that with the increasing acceptance of MR angiography, the use of catheter angiography is declining nationally. Primary prevention of stroke may be possible by transfusion therapy if patients at highest risk can be

c identified. In a landmark paper published in 1992, Adams et al. [81] reported the use of transcranial Doppler US to identify those patients with sickle cell disease at high risk of stroke [81]. An abnormal result was defined as a flow of L 170 cm/s in the middle cerebral artery, a result determined by post-hoc analysis to maximize the predictive value of the test. Some 190 patients were followed for an average of 29 months. Of these, 23 patients had an abnormal US examination and 7 clinically apparent strokes occurred, 6 among the 23 patients with an abnormal US result [81]. This study was confined to patients with homozygous sickle cell anemia. Minor hemoglobinopathies such as Hb SC disease or Hb S/beta thalassemia were excluded because of the significantly lower risk of stroke in these patients [81]. This study demonstrated that the transcranial US can identify children with sickle cell disease who are at highest risk for cerebral infarction [81]. In a well-designed study published in 1993, Seibert et al. [82] examined the sensitivity and specificity of 11 abnormalities determined by Doppler US screening. Five factors were identified that differentiated patients with

657

Fig. 15 A 4-year-old girl with a right-sided stroke. Magnetic resonance arteriogram shows high-grade narrowing of the left middle cerebral artery (arrow)

a stroke from a control patient. These were maximum velocity in the ophthalmic artery of > 35 cm/s; middle cerebral artery mean velocity of > 170 cm/s; resistive index in the ophthalmic artery < 50; velocity of the ophthalmic artery equal to or greater than the velocity in the ipsilateral cerebral artery, and velocity in the posterior cerebral, vertebral, or basilar arteries greater than that in the middle cerebral artery. Receiver operator characteristic analysis revealed that if two or more factors were present, the sensitivity was 78.3 %, the specificity was 87.1 %, and the positive predicted value was 69.2 % [82]. Siegel et al. showed that scanning the middle cerebral artery alone and using cutoffs of maximum velocity < 100 cm/s or > 200 cm/s could detect overt or silent infarctions with the sensitivity of 75 % and a specificity of 92 % [70]. In an interesting follow-up to their original study, Seibert et al. [83] published an 8-year follow-up and showed that of 4 of the original 46 control patients who had positive MRA and transcranial Doppler studies, 3 had subsequently developed clinical stroke. None of the nine original patients with positive transcranial Doppler and positive MRI but negative MRA had developed stroke. They recommended screening asymptomatic children with transcranial Doppler and, if any factors are positive, the patient should be evaluated with MRA for stenosis. If MRA is positive, hypertransfusion or other therapy to prevent stroke should be considered [83]. In July 1998, Adams et al. [69] published another landmark paper which demonstrated that transfusion

therapy, in a group of sickle cell disease children identified by transcranial Doppler US as being at high risk for stroke, could significantly reduce the risk of a first stroke. In this study, children with sickle cell anemia and no history of stroke had at least two transcranial Doppler studies that showed the time-averaged mean blood flow velocity in the internal carotid or middle cerebral artery to be at least 200 cm/s or higher. All patients were either homozygous sickle cell disease or sickle beta 0 thalassemia. A total of 130 children were identified; 63 were randomly assigned to receive blood transfusions and 67 to receive standard care [69]. There were 10 cerebral infarctions and one intracerebral hematoma in the standard-care group as compared to the 1 infarction in the transfusion group [69]. Following these results, the National Institutes of Health have issued guidelines recommending transcranial Doppler ultrasound in all children with sickle cell disease aged 2±16 years approximately every 6 months and transfusion therapy for any patient with elevated cerebral arterial blood flow of > 200 cm/s on two occasions [84]. These recommendations have not yet been followed in all institutions because calulations predicted that 40 percent of all patients with abnormal doppler studies would be free of strokes even without transfusions. There is thus a reluctance to expose 40 % of patients to unneeded long-term transfusions.

Renal disease Deaths from renal failure become increasingly common as patients with sickle cell disease live into adulthood. Deaths in childhood from renal failure are, however, rare. The primary renal diseases encountered in the pediatric population are enuresis, proteinuria, priapism, and a recently described malignant renal tumor, renal medullary carcinoma, largely confined to patients with sickle cell trait. It has long been noted that the US appearance of the kidneys in patients with sickle cell disease may be abnormal [85]. Renal length is significantly larger in sickle cell disease than in normal controls [86] and is also larger than in Hb SC disease. There is a significant positive correlation with reticulocyte count. Possible causes include glomerular hypertrophy and increased renal blood flow [86]. The echo texture may also be abnormal. A cohort study of 315 patients aged 10±20 years with sickle cell disease found 89 % of homozygous sickle cell disease patients had a normal sonographic appearance [87]. Approximately 7 % had diffusely increased echogenicity and 3 % had abnormal echogenicity confined to the renal medulla. It was postulated that diffusely increased echogenicity may represent glomerular and interstitial fibrosis and that echogenicity confined to the medulla may be the result of iron deposition. Zinn et al. [85] reported

658

11 patients with sickle hemoglobinopathies who had varying patterns of abnormally increased central echogenicity with preservation of at least a thin hypoechoic cortical rim. Primary nocturnal enuresis is a common problem affecting up to one-third of patients with sickle cell disease and related hemoglobinopathies [88]. A small series showed that 60 % of patients had complete or partial resolution of symptoms with intranasal desmopression [88]. Of patients with sickle cell disease, 15±20 % will have proteinuria [88]. Proteinuria, particularly albuminuria, is a marker of glomerular injury and a low hematocrit. In these patients this has been shown to correlate with increasing urinary albumin excretion indicating increasing glomerular injury [89]. Priapism is a condition marked by painful failure of detumescence of the penis [90]. Most children with sickle cell anemia will respond well to noninvasive medical intervention and more than half of adults will respond poorly [90]. 99mTc penile scans are useful to distinguish infarctive low-flow states from normal-flow priapism [91]. Those with low-flow states are felt to have a worse prognosis and are aggressively treated with aspiration and irrigation of the cavernosa. This pattern is usually confined to postpubertal children and adults [90]. A review of 400 pediatric male admissions to a children's hospital identified 8 patients over an 18month period. The patients were aged between 5 and 19 years [92]. At 99mTc penile scan, four had a low-flow and four had a high-flow state. Three resolved with hydration alone and five received exchange transfusions of whom three required shunt procedures. Significantly, all postpubertal patients had a severe clinical course and 99mTc penile scans were not helpful in predicting clinical course [92]. Another review from the same institution using 99mTc pertechnetate/red blood cells found no correlation between scintigraphic pattern and subsequent sexual potency [93]. A recent review article of priapism in postpubertal sickle cell disease patients concluded that, although it was a devastating complication, there was no clear consensus regarding management. Techniques such as cavernosonography and radionuclide scanning were useful to distinguish infarctive lowflow states with hypoxia and acidosis and normal-flow

priapism without acidosis, but the clinical usefulness of this information remains controversial and awaits prospective, multi-institutional trials [90]. Recently a highly aggressive tumor of the renal medulla has been described, largely confined to patients with sickle cell trait [94, 95]. Davis et al. [95] described 34 cases of renal medullary carcinoma, 1 patient with Hb SC disease, 32 patients with sickle cell trait, and 1 patient with neither. The ages range from 10 to 39 years; the dominant mass was in the renal medulla and the tumors had usually metastasized at the time of diagnosis with a duration of life after surgery of only 15 weeks. The radiographic findings were described by Davidson et al. [94] in five patients all with sickle cell trait or sickle SC disease.

Conclusion Much progress has been made in our understanding of sickle cell disease since its first published recognition nearly 90 years ago. Among the major radiologic challenges that remain are to find a method of differentiating osteomyelitis from infarctions and to identify the various etiologies presenting as ACS. The natural history of the disease has been well studied with the Jamaican cohort studies and the cooperative studies in the United States. The median survival age in the U. S. has been described to now be in the mid-40-year range. Improvements have been made in the clinical management of the illness. The prevention of sepsis in adults and children with improved bacterial vaccines and the use of prophylactic antibiotics in early childhood are examples of this improvement. The efficacy of hydroxyurea in preventing pain attacks and number of transfusions needed promises to alter favorably the natural history of the disease [96, 97]. Safer blood products and improved iron-chelation regimens have made the option of chronic transfusions to prevent recurrent strokes more feasible, although not risk-free [98]. Finally, there is the dawning of an era when definitive treatment and even cure is possible, with improvements in allogeneic bone-marrow transplantation, intrauterine stem-cell transplantation, and gene therapy [99±102].

References 1. Lane PA (1996) Sickle cell disease. Pediatr Clin North Am 43: 639±664 2. Herrick JB (1910) Peculiar elongated and sickle-shaped red blood corpuscles in a case of severe anemia. Arch Intern Med 6: 517±521

3. Pauling L, Itano HA, Singer ST, et al (1949) Sickle cell anemia, a molecular disease. Science 110: 543±546 4. Francis RB Jr, Johnson CS (1991) Vascular occlusion in sickle cell disease: current concepts and unanswered questions. Blood 77: 1405±1414

5. Hebbel RP (1991) Beyond hemoglobin polymerization: the red blood cell membrane and sickle cell disease pathophysiology. Blood 77: 214±237

659

6. Davis H, Moore RM Jr, Gergen PJ (1997) Cost of hospitalizations associated with sickle cell disease in the United States. Public Health Rep 112: 40±43 7. Schneider RG, Hightower B, Hosty TS, et al (1976) Abnormal hemoglobin in a quarter million people. Blood 48: 629±637 8. Ohene-Frempon K, Nkrumah FK (1994) Sickle cell disease in Africa. In: Embury SH, Hebble RP, Nohand AS, et al (eds). Sickle cell disease: basic principles and clinical practice. Raven Press, New York, pp 423±426 9. Salamah MM, Mallouh AA, Hamdan JA (1989) Acute splenic sequestration crises in Saudi children with sickle cell disease. Ann Trop Paediatr 9: 115±117 10. Schiliro G, Comisi FF, Testa R, et al (1990) Hematological findings in 375 Sicilians with Hb S trait. Haematologica 75: 113±116 11. Sears DA (1994) Sickle cell trait. In: Embury SH, Hebble RP, Nohand AS, et al (eds) Sickle cell disease: basic principles and clinical practice. Raven Press, New York, pp 381±394 12. Sateriale M, Hart P (1985) Unexpected death in a black military recruit with sickle cell trait: case report. Mil Med 150: 602±605 13. Castro O, Brambilla DJ, Thorington B, et al (1994) The ACS in sickle cell disease: incidence and risk factors. Blood 84: 643±649 14. Charache S, Scott JC, Charache P (1979) ªAcute chest syndromeº in adults with sickle cell anemia. Arch Intern Med 139: 67±69 15. Weil JV, Castro O, Malik AB, et al (1993) NHLBI workshop summary. Pathogenesis of lung disease in sickle hemoglobinopathies. Am Rev Respir Dis 148: 249±256 16. Vichinsky EP, Styles LA, Colangelo LH, et al (1997) Acute chest syndrome in sickle cell disease: clinical presentation and course. Cooperative study of sickle cell disease. Blood 89: 1787±1792 17. Vichinsky E, Styles L (1996) Pulmonary complications. Hematol Oncol Clin North Am 10: 1275±1287 18. Bunn HF (1997) Pathogenesis and treatment of sickle cell disease. N Engl J Med 337: 762±769 19. Martin L, Buonomo C (1997) Acute chest syndrome of sickle cell disease: radiographic and clinical analysis of 70 cases. Pediatr Radiol 27: 637±641

20. Kirkpatrick MB, Haynes J Jr, Bass JB Jr (1991) Results of bronchoscopically obtained lower airway cultures from adult sickle cell disease patients with the acute chest syndrome. Am J Med 90: 206±210 21. Poncz M, Kane E, Gill FM (1985) Acute chest syndrome in sickle cell disease: etiology and clinical correlates. J Pediatr 107: 861±866 22. Oppenheimer EH, Esterly JR (1971) Pulmonary changes in sickle cell disease. Am Rev Respir Dis 103: 858±859 23. Vichinsky EP (1994) National Acute Chest Study Group: acute chest syndrome (ACS) in sickle cell disease (SCD). Blood [Suppl] 10: 219 a 24. Bhalla M, Abboud MR, McCloud TC, et al (1993) Acute chest syndrome in sickle cell disease: CT evidence of microvascular occlusion. Radiology 187: 45±49 25. Vichinsky E, Williams R, Das M, et al (1994) Pulmonary fat embolism: a distinct cause of severe acute chest syndrome in sickle cell anemia. Blood 83: 3107-3112 26. Styles LA, Schalkwijk CG, Aarsman AJ, et al (1996) Phospholipase A2 levels in acute chest syndrome of sickle cell disease. Blood 87: 2573±2578 27. Lisbona R, Derbekyan V, NovalesDiaz JA (1997) Scintigraphic evidence of pulmonary vascular occlusion in sickle cell disease. J Nucl Med 38: 1151±1153 28. Walker BK, Ballas SK, Burka ER (1979) The diagnosis of pulmonary thromboembolism in sickle cell disease. Am J Hematol 7: 219±232 29. Castro O (1996) Systemic fat embolism and pulmonary hypertension in sickle cell disease. Hematol Oncol Clin North Am 10: 1289±1303 30. Rucknagel DL, Kalinyak KA, Gelfand MJ (1991) Rib infarcts and acute chest syndrome in sickle cell diseases. Lancet 337: 831±833 31. Sisayan R, Elgazzar AH, Webner PJ, et al (1996) Impact of bone scintigraphy on the clinical management of a patient with sickle cell anemia and recent chest pain. Clin Nucl Med 21: 523±526 32. Yam TL, Li CY (1994) The spleen. In: Embury SH, Hebble RP, Nohand AS, et al (eds) Sickle cell disease: basic principles and clinical practice. Raven Press, New York, pp 555±566 33. VanRhee F, Balsitis M, French EA (1991) Fatal splenic sequestration crisis in adult sickle cell beta thalassemia. Postgrad Med J 67: 907±908

34. Roshkow JE, Sanders LM (1990) Acute splenic sequestration crisis in two adults with sickle cell disease: US, CT and MR imaging findings. Radiology 177: 723±725 35. Powars DR (1994) Natural history of disease: the first two decades. In: Embury SH, Hebble RP, Nohand AS, et al (eds). Sickle cell disease: basic principles and clinical practice. Raven Press, New York, pp 395±412 36. King KE, Ness PM (1996) Treating anemia. Hematol Oncol Clin North Am 10: 1305±1320 37. Pappo A, Buchanan GR (1989) Acute splenic sequestration in a 2-month-old infant with sickle cell anemia. Pediatrics 84: 578±579 38. Mallouh AA, Qudah A (1993) Acute splenic sequestration together with aplastic crisis caused by human parovirus B19 in patients with sickle cell disease. J Pediatr 122: 593±595 39. Emond AM, Collis R, Darvill D, et al (1985) Acute splenic sequestration in homozygous sickle cell disease: natural history and management. J Pediatr 107: 201±206 40. Topley JM, Rogers DW, Stevens MC, et al (1981) Acute splenic sequestration and hypersplenism in the first five years in homozygous sickle cell disease. Arch Dis Child 56: 765±769 41. Powell RW, Levine GL, Wang YN, et al (1992) Acute splenic sequestration crisis in sickle cell disease: early detection and treatment. J Pediatr Surg 27: 215±218 42. Aquino VM, Norvell JM, Buchanan GR (1997) Acute splenic complications in children with sickle cell-hemoglobin C disease. J Pediatr 130: 961±965 43. Sarma PS (1989) Acute splenic sequestration crisis in a young woman with homozygous sickle cell anemia. Postgrad Med J 65: 105±107 44. Svarch E, Vilorio P, Nordet I, et al (1996) Partial splenectomy in children with sickle cell disease and repeated episodes of splenic sequestration. Hemoglobin 20: 393±400 45. Walker TM, Serjeant GR (1993) Focal echogenic lesions in the spleen in sickle cell disease. Clin Radiol 47: 114±116 46. Levin TL, Berdon WE, Haller JO (1996) Intrasplenic masses of ªpreservedº functioning splenic tissue in sickle cell disease: correlation of imaging findings (CT, ultrasound, MRI and nuclear scintigraphy). Pediatr Radiol 26: 646±649

660

47. Buchanan GR (1996) Differentiation of bone infarct from infection in a child with sickle cell disease. Pediatr Infect Dis J 15: 725 48. Burnett MW, Bass JW, Cook BA (1998) Etiology of osteomyelitis complicating sickle cell disease. Pediatrics 101: 296±297 49. Mandell GA (1996) Imaging in the diagnosis of musculoskeletal infections in children. Curr Probl Pediatr 26: 218±237 50. Rao S, Solomon N, Miller S, et al (1985) Scintigraphic differentiation of bone infarction from osteomyelitis in children with sickle cell disease. J Pediatr 107: 685±688 51. Harcke TH, Mandell GA (1994) Musculoskeletal scintigraphy. In: Miller JH, Gelfand NJ (eds) Pediatric nuclear imaging. Saunders, Philadelphia, pp 253±286 52. Amundsen TR, Siegel MJ, Siegel BA (1984) Osteomyelitis and infarction in sickle cell hemoglobinopathies: differentiation by combined technetium and gallium scintigraphy. Radiology 153: 807±812 53. Fernandez-Ulloa M, Vasavada PJ, Black RR (1989) Detection of acute osteomyelitis with indium-III labeled white blood cells in a patient with sickle cell disease. Clin Nucl Med 14: 97±100 54. Stark JE, Glasier CM, Blasier RD, et al (1991) Osteomyelitis in children with sickle cell disease: early diagnosis with contrast enhanced CT. Radiology 179: 731±733 55. Bonnerot V, Segab B, de Montalembert M, et al (1994) GadoliniumDOTA enhanced MRI of painful osseous crises in children with sickle cell anemia. Pediatr Radiol 24: 92±95 56. Mankad VN, Williams JP, Harpen MD, et al (1990) Magnetic resonance imaging of bone marrow in sickle cell disease: clinical, hematologic and pathologic correlations. Blood 75: 274±283 57. Werlin SL, Scott JP (1996) Is biliary sludge a stone-in-waiting? J Pediatr 129: 321±322 58. Okuonghae HO, Szlachetka R, Sagay AS (1993) Cholelithiasis in children with homozygous sickle cell anemia in northern Nigeria. Trop Geogr Med 45: 44±45 59. Sheehy TW (1977) Sickle cell hepatopathy. South Med J 70: 533±538 60. Betrosian A, Balla M, Kafiri G, et al (1996) Reversal of liver failure in sickle cell vaso-occlusive crisis. Am J Med Sci 311: 292±295

61. Shao SH, Orringer EP (1995) Sickle cell intrahepatic cholestasis: approach to a difficult problem. Am J Gastroenterol 90: 2048±2050 62. Lachman BS, Lazerson J, Starshak RJ, et al (1979) The prevalence of cholelithiasis in sickle cell disease as diagnosed by ultrasound and cholecystography. Pediatrics 64: 601±603 63. Sarnaik S, Slovis TL, Corbett DP, et al (1980) Incidence of cholelithiasis in sickle cell anemia using the ultrasonic grey-scale technique. J Pediatr 96: 1005±1008 64. Walker TM, Serjeant GR (1996) Biliary sludge in sickle cell disease. J Pediatr 129: 443±445 65. Winter SS, Kinney TR, Ware RE (1994) Gallbladder sludge in children with sickle cell disease. J Pediatr 125: 747±749 66. Leikin SL, Gallagher D, Kinney TR, et al (1989) Cooperative study of sickle cell disease. Mortality in children and adolescents with sickle cell disease. Pediatrics 84: 500±508 67. Balkaran B, Char G, Morris JS, et al (1992) Stroke in a cohort of patients with homozygous sickle cell disease. J Pediatr 120: 360±366 68. Lusher JM, Haghighat H, Khalifa AS (1976) A prophylactic transfusion program for children with sickle cell anemia complicated by CNS infarction. Am J Hematol 1: 265±273 69. Adams RJ, McKie VC, Hsu L, et al (1998) Prevention of a first stroke by transfusions in children with sickle cell anemia and abnormal results on transcranial Doppler ultrasonography. N Engl J Med 339: 5±11 70. Siegel MJ, Luker GD, Glauser TA (1995) Cerebral infarction in sickle cell disease: transcranial Doppler ultrasound versus neurologic examination. Radiology 197: 191±194 71. Moran CJ, Siegel MJ, DeBaun MR (1998) Sickle cell disease: imaging of cerebrovascular complications. Radiology 206: 311±321 72. Barboriak DP, Provenzale JM (1998) MR arteriography of intracranial circulation. AJR 171: 1469±1478 73. Mathews VP, Bryan RN (1987) Intracranial hemorrhage in occlusive vascular diseases. Neuroimaging Clin N Am 2: 221±233 74. Ebisu T, Tanaka C, Umeda M, et al (1997) Hemorrhagic and nonhemorrhagic stroke: diagnosis with diffusionweighted and T2-weighted echo-planar MR imaging. Radiology 203: 823±828

75. Schaefer PW, Budzik RF, Koroshetz W, et al (1997) Comparison of diffusion MR imaging, routine MR imaging and CT in hyperacute stroke. Presented at the 35 th Annual Meeting of the American Society of Neuroradiology, 18±22 May, 1997, Toronto, Ontario, Canada 76. Noguchi K, Ogawa T, Inugami A, et al (1995) Acute subarachnoid hemorrhage: MR imaging with fluid-attenuated inversion recovery pulse sequences. Radiology 196: 773±777 77. Atlas SW, Sheppard L, Goldberg HI, et al (1997) Intracranial aneurysms: detection and characterization with MR angiography with use of an advanced postprocessing technique in a blinded-reader study. Radiology 203: 807±814 78. Atlas SW (1994) MR angiography in neurological disease. Radiology 193: 1±16 79. Gillams AR, McMahon L, Weinberg G, et al (1998) MRA of the intracranial circulation in asymptomatic patients with sickle cell disease. Pediatr Radiol 28: 283±287 80. Crosby DL, Turski PA, Davis WL (1992) Magnetic resonance angiography and stroke. Neuroimaging Clin N Am 2: 509±531 81. Adams R, McKie V, Nichols F, et al (1992) The use of transcranial ultrasonography to predict stroke in sickle cell disease. N Engl J Med 326: 605±610 82. Seibert JJ, Miller SF, Kirby RS, et al (1993) Cerebrovascular disease in symptomatic and asymptomatic patients with sickle cell anemia: screening with duplex transcranial Doppler US ± correlation with MR imaging and MR angiography. Radiology 189: 457±466 83. Seibert JJ, Glasier CM, Kirby RS, et al (1998) Transcranial Doppler, MRA and MRI as a screening examination for cerebrovascular disease in patients with sickle cell anemia: an 8-year study. Pediatr Radiol 28: 138±142 84. Adams R, Brambilla D (1997) Clinical alert: periodic transfusions lower stroke risk in children with sickle cell anemia. National Heart, Lung and Blood Institute, September, p 18 85. Zinn D, Haller JO, Cohen HL (1993) Focal and diffuse increased echogenicity in the renal parenchyma in patients with sickle hemoglobinopathies ± an observation. J Ultrasound Med 12: 211±214

661

86. Walker TM, Beardsall K, Thomas PW, et al (1996) Renal length in sickle cell disease: observations from a cohort study. Clin Nephrol 46: 384±388 87. Walker TM, Serjeant GR (1995) Increased renal reflectivity in sickle cell disease: prevalence and characteristics. Clin Radiol 50: 566±569 88. Figueroa TE, Benaim E, Griggs ST, et al (1995) Enuresis in sickle cell disease. J Urol 153: 1987±1989 89. Faulkner M, Turner EA, Deus J, et al (1995) Severe anemia: a risk factor for glomerular injury. JAMA 87: 209±213 90. Powars DR, Johnson CS (1996) Priapism. Hematol Oncol Clin N Am 10: 1363±1372 91. Hashmat AI, Raju I, Singh I, et al (1989) 99mTc penile scan: an investigative modality in priapism. Urol Radiol 11: 58±60

92. Miller ST, Rao SP, Dunn EK, et al (1995) Priapism in children with sickle cell disease. J Urol 154: 844±847 93. Dunn EK, Miller ST, Macchia RJ, et al (1995) Penile scintigraphy for priapism in sickle cell disease. J Nucl Med 36: 1404±1407 94. Davidson AJ, Choyke PL, Hartman DS, et al (1995) Renal medullary carcinoma associated with sickle cell trait: radiologic findings. Radiology 195: 83±85 95. Davis CJ Jr, Mostofi FK, Sesterhenn IA (1995) Renal medullary carcinoma. The seventh sickle cell nephropathy. Am J Surg Pathol 19: 1±11 96. Charache S, Terrin ML, Moore RD, et al (1995) Effect of hydroxyurea on the frequency of painful crises in sickle cell anemia. N Engl J Med 332: 1317±1322 97. Steinberg MH (1999) Management of sickle cell disease. N Engl J Med 340: 1021±1030

98. Sheth S, Ruzal-Shapiro C, Hurlet-Jensen A, et al (1997) Pulmonary embolism developing in patients with sickle cell disease on hypertransfusion and IV deferoxamine chelation therapy. Pediatr Radiol 27: 926±928 99. Charache S (1996) Experimental therapy. Hematol Oncol Clin N Am 10: 1373±1382 100. Walters MC, Patience M, Sanders JE, et al (1995) Marrow transplantation for sickle cell disease: results of a multicenter international trial. Blood [Suppl] 86: 417 a 101. Leiden JM (1995) Gene therapy ± promise, pitfalls, and prognosis. N Engl J Med 333: 871±873 102. Walters MC, Patience M, Leisenring W (1996) Bone marrow transplantation for sickle cell disease. N Engl J Med 335: 369±376