This 11-year-old black boy with sickle cell disease (SCD) was admitted with abdominal pain of 1 day's duration. Several hours before admission, he developed nonbloody, nonbilious emesis and intermittent midepigastric abdominal pain. His medical history was remarkable for multiple pain crises and one episode of acute chest syndrome requiring transfusion. His medications include folic acid and hydroxyurea.
On examination, he was a tired-appearing boy who was lying in bed, groaning in pain. Temperature was 36.3° C, pulse 102, respiratory rate 26, blood pressure 118/66, and oxygen saturation 96% in room air. Height and weight were in the 25th percentile. There was marked scleral icterus. S1 and S2 were normal. There was a II/VI systolic ejection murmur heard best at the left lower sternal border. Lungs were clear. Abdomen was soft and distended. Bowel sounds were normal. There was epigastric and right upper quadrant tenderness to palpation with voluntary guarding without rebound tenderness or organomegaly. Neurologic examination was unremarkable.
Robert Listernick, MD, moderator: How should one follow the uncomplicated patient with SCD?
Robert Liem, MD, pediatric hematologist: All children with SCD should be followed in a comprehensive SCD clinic. We try to see each child every 6 months. First, we try to assure that all immunizations are up to date, including the 7-valent pneumococcal conjugate vaccine, the 23valent pneumococcal polysaccharide vaccine, yearly influenza immunization, the quadrivalent meningococcal polysaccharide vaccine, and the Haemophilus influenza vaccine. Blood pressure should be monitored because relatively elevated blood pressure has been associated with an increased risk of stroke. We also look carefully for the development of proliferative retinopathy, restrictive lung disease, proteinuria, cholelithiasis, and leg ulcers in older patients.
Dr. Listernick: Is their growth affected at all?
Dr. Liem: Yes. They tend to be shorter than their genetic potential and to enter puberty later.
Dr. Listernick: Do they develop central nervous system disease in the absence of overt strokes?
Dr. Liem: We have begun to perform cognitive testing routinely as these children get older. Recent research has begun to show that a subgroup of children is at risk for "silent" strokes, which affect cognitive ability.
Dr. Listernick: What about prophylactic antibiotics?
Dr. Liem: Once children are confirmed as having SCD by our state neonatal screening program, they are given penicillin prophylaxis to prevent pneumococcal infections. Antibiotics should be used at least until the fifth birthday. Continued use after this age is controversial and varies between clinics.
Dr. Listernick: Are there any predictors of disease severity?
Dr. Liem: Children with chronically high white blood cell counts and low fetal hemoglobin levels are at higher risk for recurrent vasoocclusive crises. Children with a separate disorder, hereditary persistence of fetal hemoglobin, have high hemoglobin F levels in each red blood cell and a much lower risk of vasoocclusion and other complications of this disease.
Dr. Listernick: What can be done for the child who has recurrent vasoocclusive crises?
Dr. Liem: Hydroxyurea decreases the number of crises in some children by increasing the percentage of hemoglobin F in the red blood cells. Unfortunately, it's only in a tablet form and needs to be taken daily. We can monitor compliance by following the mean corpuscular volume, which should become macrocytic. There is a theoretical increased risk of malignancy with this treatment, and some patients may become neutropenic. We're finding increasing numbers of children who are at increased risk of stroke by using transcranial Doppler imaging. We have started many on chronic transfusion therapy to decrease their risk, but we are beginning to see the effects of iron overload, necessitating chelation therapy.
Dr. Listernick: How can you tell whether this child's abdominal pain is due to a vasocclusive crisis or to a different intraabdominal pathology?
Dr. Liem: Abdominal pain may be very confusing in a child with SCD. Varying etiologies include vasocclusive crisis, cholecystitis and cholelithiasis, pancreatitis, and liver infarction.
Dr. Listernick: How can we distinguish among these diagnoses?
David Hoover, MD, pediatric surgeon: Obviously, laboratory testing will be very helpful. Children with SCD tend to have pain crises recurrently in the same location. If this child hasn't had abdominal pain crises in the past, or if the present pain is out of proportion to the pain of previous crises, one should suspect an intraabdominal process.
Dr. Liem: Certainly if there is a history of gallstones, pancreatitis and cholecystitis go higher on the list.
Dr. Listernick: Why should these children get recurrent vasoocclusive crises in the same location?
Dr. Liem: In my experience it's true, although I don't understand it. It's certainly conceivable that there's a sickle cell vasculopathy that has damaged the local endothelium and altered vascular tone, predisposing to recurrent vasooclusion locally.
Dr. Listernick: The laboratory findings were as follows: hemoglobin 9.2 g/dL, white blood cell count 26,000/mm3 with 90% neutrophils, 5% immature neutrophils, reticulocyte count 10.5%; AST 152 IU, ALT 252 IU, alkaline phosphatase 320 IU, total bilirubin 31 mg/dL, direct bilirubin 18 mg/dL, amylase 2217 IU, and lipase 2129 IU.
Karan Emerick, MD, pediatric hepatologist: Obviously, he has pancreatitis. In a patient with SCD, the first consideration would be cholelithiasis leading to common bile duct obstruction. This is supported by the markedly elevated direct bilirubin.
Dr. Listernick: How useful are the biochemical markers of pancreatitis?
Dr. Emerick: Although the serum amylase may be elevated in several conditions other than pancreatitis, an elevated serum lipase is very specific for pancreatitis.
Dr. Listernick: What is the correct imaging procedure in this situation?
Cynthia Rigsby, MD, pediatric radiologist: Ultrasonography is the procedure of choice when common bile duct obstruction from gallstones is suspected. Gallstones may be missed on computerized tomography (CT) if they are isodense compared with the surrounding fluid. Sludge is poorly visualized by CT.
Dr. Listernick: How well can you image the pancreas by ultrasonography?
Dr. Rigsby: The sensitivity of ultrasonography for pancreatic inflammation is low. If one is looking for a complication of pancreatitis such as pseudocyst formation, CT is far superior. Ultrasonography may miss retroperitoneal fluid collections that may develop as a consequence of pancreatitis.
Dr. Emerick: For the clinician, ultrasound will give us a better look at the ducts, the material within the ducts, and the gall bladder, including its wall thickness.
Dr. Rigsby: This patient's ultrasound demonstrates that the common bile duct is dilated. Although we did not see a distal common duct stone, it can be easily missed by ultrasonography. The gall bladder has multiple echogenic foci within sludge, indicating multiple stones; the gallbladder wall is not thickened. The pancreas is large and edematous. A CT scan was performed several days later.
Dr. Listernick: Why did he have the CT scan?
Dr. Hoover: His bilirubin continued to climb, and he remained febrile. We were concerned that he had developed an infected, necrotic pancreas. If we had found any complex fluid collections and he remained febrile, we would have considered asking the interventional radiologist to needle the fluid to identify pathogenic bacteria.
Dr. Listernick: Isn't fever commonly seen in pancreatitis?
Dr. Hoover: Yes, but it usually resolves after several days.
Dr. Rigsby: The CT scan showed bilateral pleural effusions and lower lobe infiltrates, common complications of pancreatitis. There was some edema around the pancreas but no evidence of fluid collections or necrosis.
Dr. Liem: This child was known to have gallstones for several years but the mother hadn't wanted surgery. We don't routinely screen asymptomatic SCD patients for gallstones even though the incidence is between 30% and 50% as children get older.
Dr. Listernick: Is the presence of gallstones sufficient to warrant cholecystectomy?
Dr. Hoover: I'd say yes, even if the patient were asymptomatic. We would like to avoid the complications of gallstones, primarily pancreatitis. We would much prefer to operate on these children electively, when they are healthy.
Dr. Listernick: Should every child who has gallstones undergo cholecystectomy?
Dr. Emerick: If a child has a known predisposition to the formation of gallstones, such as a congenital hemolytic anemia or a bile salt transporter defect, cholecystectomy is indicated, as gallstones will continue to form. However, if a child has a single gallstone associated with a temporary precipitant, such as the use of total parenteral nutrition, you might use ursodeoxycholic acid to help dissolve the stone in the hope that it won't recur.
Dr. Listernick: What's the treatment for acute pancreatitis?
Dr. Emerick: Generally, we use bowel rest, intravenous hydration, and pain medication. Continuous nasogastric suction was the standard of care 20 years ago but has been eliminated from the regimen unless the patient has intractable vomiting.
Dr. Listernick: This patient was receiving antibiotics because he was febrile and had SCD. Are antibiotics indicated for the routine patient with pancreatitis?
Dr. Hoover: Antibiotics aren't indicated for uncomplicated pancreatitis. However, if the patient has evidence of biliary obstruction and fever, antibiotics should be used for presumed cholecystitis. In addition, we often use Ranson's criteria, a list of prognostic criteria, to decide on the need for further intervention or antibiotics. Poor prognostic factors include an elevated white blood cell count, hyperglycemia, acidosis, and elevated serum transaminases.
Dr. Listernick: In looking at his chart, I saw that there was discussion about possibly performing endoscopic retrograde cholangiopancreatography (ERCP). What are the indications for its use?
Dr. Hoover: His bilirubin was rising during the first several days of hospitalization. Our concern was that there was a common bile duct stone causing obstruction and worsening pancreatitis. ERCP is the procedure of choice for removing such stones. Ultimately, he began to improve, obviating the need for ERCP.
Judah Jona, MD, pediatric surgeon: I want to emphasize that ERCP should be used cautiously in patients with pancreatitis, as it may exacerbate the inflammatory process. In addition, perhaps the most important function of ERCP is the performance of a sphincterotomy, which would help to decompress the biliary system.
Dr. Listernick: I assume that this patient needs a cholecystectomy at some point.
Dr. Hoover: He had one of the worst complications of gallstones, severe pancreatitis. Most surgeons would wait several weeks to perform the operation in order to let the pancreas "cool down," recognizing that there's a risk that he would pass another stone. Most pediatric surgeons would perform the procedure laparoscopically.
Dr. Listernick: How would you prepare this child for surgery?
Dr. Liem: In the past, we would have used either exchange transfusions or serial transfusions to reduce his level of hemoglobin S to less than 30%. Recent studies have shown that a simple transfusion that raises the hemoglobin level to approximately 10 g/dL is equally effective and safe when compared with a more aggressive transfusion protocol.
Stephen Almond, MD, pediatric surgeon: Does this mean that the anesthesiologists' demand for a preoperative SCD screen of all black children is not necessary if their hemoglobin is above 10 g/dL?
Dr. Liem: If the only reason is to look for perioperative complications, then it isn't necessary. Of course, there may be other reasons we would need to assure ourselves that such children don't have SCD, particularly if they were born in states that don't have neonatal hemoglobinopathy screening programs.
Dr. Listernick: You shouldn't even assume that a child who was born in a state such as Illinois, which does have a program, is normal. We reported 10 children who had abnormal neonatal screens for hemoglobinopathies; none of the families had been properly notified, and all of the children presented later in life with what would have been preventable complications of their disease. Factors in the failure to notify the families included incorrect addresses and the lack of identified primary care physicians (JAMA. 1992;267(8):1095-1099).
Thank you, everybody.