This 9-year-old girl was admitted to the hospital for evaluation of abdominal pain of 5 days' duration. The pain was described as periumbilical and cramping, waxing and waning in intensity. There had been associated nausea with decreased appetite, but no fever or diarrhea. The night prior to admission, she had two episodes of nonbilious, nonbloody emesis. There was no history of cough, pharyngitis, dysuria, or other symptoms.
Her history was remarkable for three sets of myringotomy tubes and a tonsillectomy with adenoidectomy.
She had recently traveled to British Columbia and Alaska but had not been camping. The family history was unremarkable in detail.
On exam, she was an alert, healthy-appearing girl. She was afebrile, pulse 68, respiratory rate 18, blood pressure 115/60. Weight and height were both in the 25th percentile, and she had been growing and gaining weight normally. HEENT examination was unremarkable. Neck was supple. There was bilateral anterior cervical lymphadenopathy. Lungs were clear. Sl and S2 were normal without murmurs. There was periumbilical tenderness without rebound tenderness or guarding. The abdomen was soft without masses or organomegaly. Neurologic exam was unremarkable. There were no rashes.
On laboratory evaluation, hemoglobin was 13.4 g/dL, white blood cell count 11,800/mm3 with 81% neutrophils, 13% lymphocytes, 6% monocytes; platelet count 319,000/ mm3. Serum chemistries, lipase, amylase, and urinalysis were normal.
Robert Listernick, MD, moderator: How do we make sure that she doesn't have appendicitis?
Jennifer Trainor, MD, pediatrie emergency medicine physician: There's a classic history of uncomplicated appendicitis. Initially, the child has anorexia, nausea, and vague periumbilical abdominal pain. During the next 24 to 36 hours, the pain migrates to McBurney's point and intensifies, leading to focal guarding and rebound tenderness.
Often, there is a low grade fever and a mildly elevated white blood cell count with a left shift. The 5 -day history in this child and her relatively benign abdominal exam are against appendicitis.
Dr. Listernick: How does a surgeon decide whether someone has appendicitis?
Tony Chin, MD, pediatrie surgeon: If we're not sure but suspect appendicitis, we will observe the patient for 12 to 24 hours and see the evolution of symptoms. We won't give any pain medications so that we can better appreciate the development of focal peritoneal symptoms.
Dr. Listernick: What's the role of radiographie studies in evaluating children for possible appendicitis?
Sri Pillai, MD, pediatrie surgeon: It's dependent on the surgeon. If the clinical history and exam are consistent with the diagnosis, I'll operate. If not, I'll often order a computed tomography (CT) scan of the abdomen.
Mary Wyers, MD, pediatrie radiologist: The findings of appendicitis on CT scan include an enlarged appendix greater than 6 mm in width, inflammatory stranding around the appendix, and an appendiceal lumen that doesn't fill with contrast The sensitivity and specificity for these findings are both very high, above 90%. Of course, these findings are both operator and interpreter dependent Dr. Listemick: She was observed for 36 hours. The pain waxed and waned, but the nausea and vomiting stopped, and there were never any physical findings suggestive of appendicitis. She was discharged home to be followed by her pediatrician. What was the discharge diagnosis?
Dr. Chin: I believe that it was gastroenteritis.
Jeff Brown, MD, pediatrie gastroenterologist: One should never make the diagnosis of gastroenteritis without the presence of diarrhea. Certainly in infections such as rotavirus gastroenteritis, you might see vomiting without diarrhea in the first 36 hours of illness. However, there are too many more serious illnesses that can cause vomiting alone. The danger in assigning the diagnosis of "gastroenteritis" is that you might become conceptually blocked and miss the more serious condition. Five days of vomiting without diarrhea is not gastroenteritis.
Dr. Listemick: During the next week, she continued to have intermittent episodes of abdominal pain and emesis. She was scheduled to see a gastroenterologist but returned to the emergency department before the appointment. What now?
Dr.Brown: The presence of fever, weight loss, or vomiting are all factors which point to organic disease. If I had seen her after 2 weeks of episodic pain and vomiting, assuming there was nothing diagnostic in the laboratory and radiographie evaluation and the stool tested negative for occult blood, I would probably perform upper endoscopy. All of her symptoms point to a process in the upper gastrointestinal tract, perhaps allergic or eosinophilic esophagitis.
Dr. Listemick: What would be your diagnostic approach to someone whom you thought had functional abdominal pain?
Dr. Brown: First, you need to assure both yourself and the parents that there is no organic disease. At times, unfortunately, this may require endoscopy. Once you are confident of the diagnosis, you should give a straightforward explanation to the parents.
Dr. Listemick: On examination at the time of readmission, she was lying quietly in bed, but she was clearly in pain. Vital signs included temperature 37° C, pulse 100, respiratory rate 20, and blood pressure 100/60. Abdomen was soft with pain to palpation in the periumbilical area without rebound tenderness. There was no guarding. In addition, there were at most 10 2-mm purplish papules and several petechiae on the lower extremities and buttocks. On questioning, her mother stated that this rash had been present on the previous admission. As subtle as the rash was, there was no question in my mind that she had HenochSchonlein purpura (HSP).
Anthony Mancini, MD, pediatrie dermatologist: Nonblanching purplish papules are pathognomonic of vasculitis, usually a small vessel leukocytoclastic vasculitis. The most common type of small vessel vasculitis in children is HSP. Large vessel vasculitis includes entities such as Takayasu's arteritis, while medium-sized vessel vasculitis would include diagnoses such as Kawasaki disease and polyarteritis nodosa. Interestingly, palpable purpura rarely occurs in patients with Kawasaki disease.
Dr. Listemick: Why are the lesions of HSP predominantly in the lower extremities?
Dr. Mancini: The rash of all vasculitis syndromes predominates on the lower extremities and the buttocks, most likely due to leakage of red blood cells secondary to dependency and gravity.
Dr. Listemick: What are the clinical manifestations of HSP?
Kelly Rouster-Stevens, MD, pediatrie rheumatologist: HSP is a small vessel vasculitis that predominantly involves four organs - skin, joints, kidneys, and the intestines. A nondeforming arthritis occurs in as many as 80% of the patients generally involving the knees and ankles, and to a lesser degree, the wrists and elbows. Although there are no wellperformed controlled studies, anecdotally the arthritis almost always responds to oral corticosteroids.
Kidney disease develops in 20% to 40% of patients, most commonly microscopic hematuria. Rarely, a more severe form of rapidly progressive glomerulonephritis occurs characterized by proteinuria and nephritic syndrome. Finally, colicky abdominal pain and hematochezia are quite common, often preceding the onset of rash by as long as 2 weeks, as in this patient. Heo-ileal intussusception - as opposed to ileocolic intussusception, which occurs in non-HSP patients - may lead to bowel perforation if unrecognized. Corticosteroids will also alleviate the abdominal pain; once again, that's anecdotal data.
Dr. Listemick: What range of corticosteroid dosages do you use to treat HSP?
Dr. Rouster-Stevens: In uncomplicated outpatient cases, I may use 2 to 4 mg/kg/day of prednisone. If patients are hospitalized, I will give the steroids intravenously because oral medications may not be absorbed well in patients with gastrointestinal vasculitis. Once again, this treatment is unproven, although generally accepted as efficacious.
Dr. Wyers: The morning of admission, an abdominal ultrasound demonstrated what was felt to be an ileo-ileal intussusception. Findings on ultrasonography diagnostic of an intussusception include a "doughnut" sign consisting of a thick hypoechoic rim with a dense center and the "pseudokidney" sign on longitudinal view with a hypoechoic rim and hyperechoic center, which looks like a kidney.
Dr. Listernick: So she has an ileo-ileal intussusception, which can't be reduced by barium enema because of the presence of the ileocecal valve and the inability to generate enough hydrostatic pressure by enema to reduce it Doesn't she need immediate surgery?
Dr. Pillai: Not necessarily. She had no fever, a normal white blood cell count, and a relatively unimpressive abdominal exam without any peritoneal signs. There's no evidence that she's obstructed. In addition, we and the radiologists know that ileo-ileal intussusceptions are common radiographie findings in children who are undergoing ultrasonography or CT scans for reasons unrelated to abdominal pain. They may appear and disappear without causing any symptoms. We even see them incidentally during surgery.
Dr. Listemick: But this is a child with HSP who has a real reason to have an intussusception with bowel wall purpura producing a lead point for the intussusception. Shouldn't we be afraid that this untreated intussusception will lead to bowel infarction or a perforation?
Dr. Pillai: Her examination was benign. We felt that we could start medical therapy with corticosteroids and observe her closely for progression for the next 12 to 24 hours. In addition, a follow-up plain film of the abdomen clearly showed that the oral contrast that she had received for CT scan had bypassed the intussusception, indicating that there was no obstruction.
Dr. Wyers: I should also say that there was some disagreement among the radiologists whether this finding was due to an intussusception or simply a thickened segment of vasculitic bowel wall.
Dr. Listernick: These discussions occurred the morning of admission, and it was decided to give her prednisone 4 mg/kg/day in divided doses and observe her closely. I have to admit that I was fairly uncomfortable with this plan. However, within 3 hours of the first dose, her mother felt that her daughter was miraculously back to normal, the first time in 12 days. She was pain-free and eating normally. The next morning an abdominal ultrasound was performed, which was normal, and she was discharged to receive tapering doses of oral corticosteroids during the next 3 weeks. Dr. Pillai was absolutely right, although I don't see how the steroids could have plied their magic in just 3 hours.
She was well for 3 days but then began experiencing the same colicky abdominal pain. Her parents decided to bring her to the emergency department and on the way she had one bloody emesis. On arrival, she had several large bloody stools that were essentially all blood, estimated at 600 to 800 cc of blood. Her initial hemoglobin was no different from her discharge hemoglobin. She appeared ill but was not in shock, and she was admitted to the intensive care unit. I believe that the emergency department physicians called both surgery and gastroenterology to see if she needed endoscopy to look for the source of bleeding.
Dr. Brown: Large gastrointestinal hemorrhages are certainly seen in HSP. Her vital signs were unremarkable. It was extremely unlikely that endoscopy would reveal any treatable lesion; more likely there would be multiple areas of bowel wall purpura. As long as her abdominal exam didn't reveal any peritoneal signs indicating bowel wall necrosis and perforation, I would have restarted the steroids. In addition, the thought of performing endoscopy on very friable, vasculitic bowel was not especially appealing.
Dr. Listernick: Could she have steroid-induced gastritis and bleeding from such a short course of steroids?
Dr. Rouster-Stevens: Very unlikely. More likely, the steroids weren't getting absorbed and the vasculitis returned unchecked.
Dr. Pillai: We were concerned that the intussusception had recurred; however, an ultrasound was normal. We fluid resuscitated her and consulted the immunologists.
Dr. Rouster-Stevens: My colleague felt that she had uncontrolled vasculitis and decided to give her "pulse steroids"; she received l g of methylprednisolone daily for several days in an attempt to get the vasculitis under control.
Dr. Listernick: Her bleeding resolved within 1 day; she never required a transfusion. She was discharged after 4 days on a very slow taper of prednisone. HSP may have frequent recurrences in the first several months after presentation, so we decided to wean her very slowly. As of several weeks later, she continues to do well. Thank you, everybody.