Prolonged Fever in a 3-Year-Old With Sickle Cell Disease

A 3-year-old male with a significant past medical history of Hb SS disease and eczema presented to the emergency department (ED) during the spring season with fever and a rash. The highest recorded temperature at home was 39.3°C. The rash was nonpruritic, noted the day before, and initially appeared on both hands and subsequently spread to cover his entire body. His mother reported that he appeared fatigued, with reduced activity and poor oral intake. At the time of presentation, there were no complaints of cough, runny nose, ear pain or discharge, eye redness, joint swelling, irritability, change in mental status, vomiting, or diarrhea. His home medications included penicillin V potassium and folic acid. His immunizations were delayed because he was due for pneumococcal polysaccharide vaccine and meningitis ACWY vaccine. He had not traveled outside the country recently.

In the ED, his vital signs were as follows: Temperature 37.8°C, heart rate 142 beats per minute, blood pressure 101/66 mmHg, respiratory rate 22 per minute and oxygen saturation 100% on room air. His weight and height were 14.8 kg (35th percentile) and 103 cm (81st percentile), respectively. Physical examination showed a miserable child with a faint papular, nonerythematous rash on the face and extremities, along with posterior oropharyngeal erythema. There was no enlargement of cervical nodes, conjunctival redness, or extremity swelling. His lung had clear breath sounds. A grade II/VI systolic flow murmur was detected, and his spleen was palpable 5 cm below the costal margin. In the ED, blood samples were collected, and he was placed on ceftriaxone. The patient was admitted to the pediatric floor for continuing IV antibiotics and monitoring. Laboratory results showed anemia, leukocytosis, and mild hyperbilirubinemia (Table 1). The child was not an established patient in our hospital, and on the basis of information provided by his mother, his baseline Hb ranged between 9 and 9.5 g/dL.

What diagnoses are you considering with the above presentation, Dr Vythinathan?

The 3-year-old child with Hb SS disease presented with symptoms including fever, rash, and an enlarged spleen, accompanied by a decrease in Hb levels. Our initial differential was sepsis or a bacterial infection, given the fact that he was febrile. Other considerations included splenic sequestration crisis, although a viral infection can present similarly. At the time of admission, the rash was faint. I will consider performing a sequential abdominal examination while monitoring the trends in Hb and hematocrit.

How does a splenic sequestration typically present and what else is on your differential, Dr Basak?

Splenic sequestration typically exhibits a decrease in Hb from the patient’s usual baseline, an acutely enlarged spleen, elevated reticulocyte count, and is sometimes accompanied by mild to moderate thrombocytopenia.²  In this case, the patient presented with splenomegaly that was enlarged compared with his usual size, a reduction in Hb by about 1.5 g/dL from his baseline, and an increased reticulocyte count to 13.65%, but no evidence of thrombocytopenia. There were no records of any recent transfusions. Hypersplenism is a potential differential diagnosis, but it is a chronic process that is caused by an increased splenic phagocytic activity. This condition is associated with laboratory findings of anemia, leukopenia, and thrombocytopenia. Aplastic anemia could also result in a sudden drop in Hb levels; however, it is usually associated with a low reticulocyte count indicating marrow suppression and absence of splenomegaly. This patient did not present with abdominal pain, which could be indicative of alternative differential diagnoses for other variable pathologies.

On admission to the pediatric floor, the patient’s vital signs were as follows: Temperature 38.5°C, heart rate 156 beats per minute, blood pressure 97/64 mm Hg, and respiratory rate 34 breaths per minute. The physical examination was notable for increased work of breathing and tachycardia, along with hepatomegaly measured at 3 cm below the costal margin and splenomegaly at 5 cm below the costal margin. The patient continued receiving IV Ceftriaxone and maintenance fluids. A subsequent blood count performed after 10 hours showed severe anemia, with an Hb of 6.7 g/dL (nearly 2.5 g/dL below his baseline), and a reticulocyte count of 11.7%. The presence of symptomatic anemia prompted a transfusion of 5 mL/kg of packed red blood cells (RBCs). On the following day, the child was noted to have scleral icterus. The abdominal examination was notable for right upper quadrant tenderness and hepatomegaly (with the liver border palpable 3 cm below the right costal margin at the midclavicular line). However, the spleen size had decreased posttransfusion (with the splenic border palpable 2 cm below the left costal margin at the midclavicular line). A repeat comprehensive metabolic panel revealed a total bilirubin level of 14.9 mg/dL with a direct component of 14 mg/dL. The liver enzymes were elevated with aspartate aminotransferase (AST) at 121 U/L, alanine aminotransferase (ALT) at 71 U/L, and gamma-glutamyl transpeptidase at 170 U/L (normal range 8–61 U/L). Additional results included a creatinine level of 0.17 mg/dL, blood urea nitrogen at 3 mg/dL, albumin at 4.1 gm/dL, and a prolonged prothrombin time of 19.1 seconds (control 13–18 seconds). Urinalysis showed an RBC count of 61 per high-power field (HPF), a white blood cell (WBC) count of 1 per HPF, and significant levels of bilirubin and urobilinogen. Both urine and blood cultures were negative. A chest x-ray done showed prominent bronchovascular markings.

What are the causes of cholestasis in patients with SCD, Dr Desai?

Sickle cell hepatopathy can be broadly divided into acute and chronic causes. Acute causes include hepatic crisis, hepatic sequestration, intrahepatic cholestasis, and liver failure. Chronic manifestations include gallstones, cholangiopathy, viral hepatitis, and transfusion overload.³ 

Given the sudden change in his clinical status, this scenario leans toward an acute presentation of sickle cell hepatopathy. Hepatic crisis constitutes a type of vaso-occlusive crisis that occurs from Kupffer cell enlargement, leading to obstruction within the hepatic sinusoids. In the majority of liver diseases, ALT activity is higher than that of AST, except for conditions such as alcoholic hepatitis and Reye's syndrome. ALT is considered as a more specific biomarker of hepatocellular injury. In our patient, the higher AST elevation compare with ALT is probably associated with the hemolytic tendencies seen in SCD. Moreover, there is significant evidence linking vaso-occlusive crisis to an increase in hemolysis.⁴ 

Hepatic sequestration represents a more severe complication caused by the destruction of RBCs in the reticuloendothelial system, commonly in the spleen, but is less common in the liver.⁵  The patient had right upper quadrant pain, increase in liver size, and a drop of Hb, which could be because of hepatic sequestration. Although intrahepatic cholestasis is not a common presentation, it tends to have a more severe clinical presentation, progressing rapidly to liver failure.

On the third day of hospitalization, an ultrasound of the abdomen was done and showed an enlarged liver with severe distention of the gallbladder. There was no evidence of cholelithiasis, gallbladder wall thickening, or dilation of the common bile duct. However, the splenic sequestration was improving as evident by the regression of the splenic size, with a stable Hb level of 8.7 g/dL. The clinical picture was consistent with acute sickle cell hepatopathy.

During the hospital course, the patient continued to be febrile with maximum temperature of 40.9°C. We continued IV Ceftriaxone for empirical coverage for bacteremia in a child with Hb SS disease. There was a noticeable decrease in jaundice, with total and direct bilirubin levels trending downward from peak levels. Blood cultures, respiratory viral panel, and group A Streptococcus polymerase chain reaction, along with an extensive workup for viral hepatitis, cytomegalovirus, parvovirus, and Epstein-Barr virus were all negative.

On the fifth day of hospitalization, the patient’s Hb levels dropped to 6.7 g/dL, with a high C-reactive protein (CRP) of 12.2 mg/dL. He remained tachycardic even during periods of defervescence, which prompted a second packed RBC transfusion. His spleen and liver sizes remained stable after this period, and the liver enzymes and bilirubin levels continued to decline. With daily persistent high fevers, there was a need for additional investigations, because the preliminary cultures did not show any growth. His WBC count was elevated to 20.000/mm³, with an increased absolute eosinophil count of 2900/mm³. An echocardiogram, done on the fifth day to exclude infective endocarditis, showed a structurally normal heart with normal systolic function. There was no evidence of coronary ectasia, dilation, or aneurysm. The infectious disease team was consulted, and on the sixth day, they recommended discontinuing IV Ceftriaxone, with close monitoring of the fever curve. The patient’s general condition improved but for the high-grade fever of 40°C. On the eighth day since admission, a desquamating rash was noted on the abdomen.

What are the causes of fever of unknown origin in a 3-year-old child, Dr Kohlhoff? What were potential differentials at this point, particularly in sickle cell patients?

Evaluation of a young child for fever with no identified origin begins with a thorough history to explore the epidemiology, including any unusual exposures to infectious agents that may have been missed during the initial history taking. Any specific tests for infectious etiologies should be targeted according to exposures and risk factors. In this case, the underlying chronic medical condition of SCD was guiding the workup. Given the increased risk for invasive infections, particularly bacteremia and subsequent complications, it was critical to send enough blood cultures and repeat the physical examination for bacteremia complications. Other bacterial infections to consider in a 3-year-old who may be limited in expressing symptoms include urinary tract infection and bone and joint infections. Physical examinations and appropriate laboratory tests were done to rule out those diagnoses. The rash was not a typical “sandpaper-like” rash expected with scarlet fever; and even though the patient had a low risk of developing group A Streptococcus disease because of their penicillin V potassium prophylaxis, a throat swab was collected in the ED, and the results came back negative. The abdominal symptoms prompt evaluation for loculation or occult abscess.

On the ninth day of hospitalization, the child continued having high-grade fevers. A computed tomography of the abdomen and pelvis with and without contrast was ordered to rule out occult abscess and revealed a significant distension of the gallbladder, or hydrops, and no intraabdominal abscess was observed (Fig 1). Throughout the hospital course, sequential blood cultures and respiratory viral panels were negative. Basic screen for autoimmune conditions showed negative antinuclear antibody and rheumatoid factor. A peripheral blood smear showed only a small number of atypical lymphocytes with no evidence of blasts.

Drs Basak and Kohlhoff, how does the child’s presentation with 9 days of fever, desquamating rash, high inflammatory markers, and hydrops of the gallbladder lead to the possible differential diagnosis for the underlying cause in this case?

The finding of hydrops of the gallbladder in the absence of stones, or obstruction or cholecystitis together with prolonged high fever and transaminitis suggests incomplete Kawasaki disease (iKD) as a possible diagnosis. Because serious bacterial infections were ruled out, it may be worthwhile to administer IV immunoglobulin (IVIG) for iKD.

The child received acetylsalicylic acid (ASA) and IVIG infusion at the dose 2 gm/kg over 12 hours without any adverse reactions.

How does an iKD fit into the differential in this patient, Dr Dubey?

KD is a clinical diagnosis with echocardiography as an important adjunct. However, if incomplete KD is suspected, echocardiogram can aid in diagnosis. For diagnosing iKD, children should have fever >5 days with 2 or 3 criteria, but in infants with unexplained fever >7 days, it should be suspected. We evaluate this with other laboratory and echocardiographic findings. Echocardiogram is considered positive if any of the 3 conditions are met: Z score of left anterior descending coronary artery or right coronary artery ≥2.5, if coronary artery aneurysm is observed, or ≥3 of the other suggestive findings exist, including decreased left ventricular function, mitral regurgitation, pericardial effusion, or z scores in left anterior descending coronary artery or right coronary artery of 2 to 2.5. The echocardiogram was normal for the patient. However, a normal echocardiogram does not rule out the diagnosis of KD. Given the associated clinical diagnostic and laboratory findings, it was agreed on initiating treatment of incomplete KD with ASA and IVIG.^6 ,7  The patient did not received corticosteroids as adjunctive therapy because he was not identified as being at high risk for the development of coronary artery aneurysms.

What is the significance of eosinophilia in this patient, Dr Basak?

Causes of peripheral blood eosinophilia include helminthic infection, atopic disease, or Churg-Strauss syndrome. In a few studies, peripheral blood eosinophilia was noted, particularly in the acute stages of KD. The exact mechanism of eosinophilia in KD is unknown but may be associated with an underlying allergic or immune response. Peripheral blood eosinophilia in KD may be associated with higher incidence of coronary micro vessel lesions. Peripheral blood eosinophils were markedly increased in the acute stages and return to normal within 3 weeks of IVIG treatment.^8 ,9 

After consulting with infectious disease, hematology, and cardiology teams, we considered the possibility of incomplete KD. He received ASA and IVIG, and remained afebrile for the remainder of the hospital stay other than a minor temperature spike shortly after the IVIG administration fever curve (Fig 2). He was discharged on the 11th day of hospitalization with the WBC counts and CRP showing a downward trend (Table 2). Three days after discharge, the patient attended a follow-up appointment at the clinic. He was afebrile and well-appearing, with skin desquamation on both soles.

Assessing a patient with SCD promptly when a fever emerges is of the highest importance for several critical reasons. SCD is associated with an impaired splenic function. Hyposplenic and asplenic individuals lack immunoglobulin M memory B cells, a subset of B lymphocytes responsible for initiating a rapid, specific response to encapsulated bacteria, indicating the spleen’s involvement in their production or function. Consequently, SCD patients are more susceptible to severe bacterial infections, particularly those caused by Streptococcus pneumoniae and Haemophilus influenzae. These potentially life-threatening infections include pneumonia, sepsis, and meningitis. Furthermore, infections can trigger or exacerbate vaso-occlusive crises and acute chest syndrome.¹⁰  Therefore, an early evaluation can aid in distinguishing between these life-threatening conditions.

This case highlights the importance of considering a broad range of differentials in a child with SCD who presents with prolonged fever. Although bacterial infections are common in children with Hb SS, it is imperative to consider other causes, because fever can also herald other complications of SCD, such as aplastic crisis or acute chest syndrome.¹  It is particularly important to identify additional signs to provide clues to the diagnosis, and further laboratory investigations may be necessary.

KD is a multisystem, vasculitic condition that can lead to severe complications such as coronary artery disease and heart failure. In North America, KD is more prevalent during winter and early spring, with an estimated incidence of 25 cases per 100 000 children <5 years of age per year.⁶  The males to female ratio is 1.5:1.¹¹  The diagnosis is clinical and the criteria are fever of at least 38.3°C for a minimum of 5 days, along with at least 4 of the following signs: Bilateral nonexudative conjunctival injection; erythema of the oropharynx accompanied by red, cracked lips and a “strawberry-like” tongue; swelling or redness of the hands and feet; and polymorphous skin rash and a nonsuppurative unilateral cervical lymph node enlargement measuring >1.5 cm in diameter. On the other hand, iKD should be suspected by a fever lasting for at least 5 days, meeting 2 or 3 of the classic criteria, with elevated levels of CRP at a minimum of 3.0 mg/dL and/or an erythrocyte sedimentation rate of at least 40 mm per hour. Additionally, it requires fulfilling at least 3 supplemental laboratory criteria: Albumin level of ≤3.0 g/dL, anemia for age, elevation of alanine aminotransferase, platelet count of ≥450 000 per mm3 after 7 days, WBC count of ≥15 000 per mm3, and presence of ≥10 WBC per HPF in urinalysis.¹²  Patients with iKD, especially those who do not have ocular or oral mucosal changes, may have significant delay in diagnosis.⁶ 

Our patient exhibited a sandpaper-like rash, which was not apparent initially, but had desquamation of skin on both the soles after discharge. He did not present with conjunctival injection, lymphadenopathy, or swelling of the hands and feet. Considering 9 days of high fever, in addition to elevated inflammatory markers, worsened anemia requiring 2 packed RBC transfusions and a WBC count elevated to ≥15 000 per mm3, unexplained eosinophilia, and a hydropic gallbladder, along with the absence of plausible infectious or autoimmune markers, we decided to treat for incomplete KD.

Hydrops of gallbladder had been reported in 4% to 13.7% of patients with KD. However, it is possible that the actual incidence might be underestimated, because some patients may not have any symptoms.¹³ 

The diagnosis of KD was particularly challenging in this case because of the patient’s medical history of SCD, a rash not apparent, and the associated splenic sequestration and hepatopathy. Furthermore, because of the concurrent episode of splenic sequestration and acute hepatopathy, some laboratory criteria were less reliable, such as decreasing Hb, elevated inflammatory markers, and aminotransferases. Conversely, the association of SCD and KD is rare, with only 1 published case series present, describing 2 patients.¹⁴ 

Currently, our patient is doing well. Subsequent follow-ups revealed no recurrence of fever and normal repeat echocardiogram results.

ALT

alanine aminotransferase

ASA

acetylsalicylic acid

AST

aspartate aminotransferase

CRP

C-reactive protein

ED

emergency department

Hb SS

hemoglobin variant SS

HPF

high-power field

iKD

incomplete Kawasaki disease

IV

intravenous

IVIG

intravenous immunoglobulin

RBC

red blood cell

SCD

sickle cell disease

WBC

white blood cell