A 13-year-old girl was evaluated following 5 weeks of nonspecific abdominal and back pain. The abdominal pain was described as dull and constant in both the left lower quadrant and left flank. There was no history of vomiting, diarrhea, fever, or other associated symptoms.
Several weeks later, she started complaining of mild intermittent cervical and midthoracic spine pain without any limitation of function. She was seen at an outside hospital by a pediatrie gastroenterologist who performed an abdominal computed tomographic (CT) scan as well as upper gastrointestinal endoscopy and colonoscopy; all the tests were normal including biopsies of the duodenum and colon.
In addition, the following tests were all normal: 24-hour urine for coproporphyrins, antinudear antibodies, chest x-ray, abdominal ultrasound, gallium scan, bone scan, barium enema, upper GI with small bowel follow through, and magnetic resonance imaging (MRI) of the spine. Hie only remarkable laboratory finding during this time was the complete blood cell count (CBC): hemoglobin 10.3 g/dL, mean corpuscular volume (MCV) 67, mean corpuscular hemoglobin (MCH) 20, and erythrocyte sedimentation rate (ESR) 100 mm /hour. Despite adequate iron replacement therapy, her CBC did not change over 5 weeks and the reticulocyte count remained under 1%. BoUS a pediatrie hematologist and rheumatologist felt that "the diagnosis was not in their areas."
Of note, her primary care physician had performed a tuberculin skin test during these 5 weeks; it reacted with 22mm of induration. The previous tuberculin test was 2 years earlier and had been negative.
Her review of systems was remarkable for constant fatigue and occasional night sweats. She had lost 7 pounds over the last 2 months. There was no history of rashes, fever, vomiting, diarrhea, or arthritis. Her past medical history and family history were unremarkable.
Her travel history was of interest in that 4 months prior to this illness she had been in the Black Hills where she had camped and fed both mules and prairie dogs. She was not aware of any tick bites.
On examination, she was an alert, healthy-appearing white girl. Her weight and height were both in the 25* percentile. Pertinent findings were limited to her abdomen and back. Her abdomen was soft with only minimal tenderness to palpation without any masses or organomegaly. Her rectal examination was normal; the stool was Hemoccult® negative. She had very mild pain to palpation over her midcervical and thoracic spine without any paraspinal muscular tenderness. Her back was straight and had full range of motion. She was Tanner 2. The remainder of the examination was unremarkable.
Robert Listernick, MD, moderator First, let's look at the evaluation she had had to date. All of her physicians thought that the most likely diagnosis was inflammatory bowel disease (IBD) because she had abdominal pain, anemia, and an elevated erythrocyte sedimentation rate. How should the primary care physician proceed when IBD is suspected?
Karan Emerick, MD, pediatric gastroenterologist: I would certainly start with a CBC, ESR, and set of chemistries including liver transaminases and a serum albumen. I would also perform stool cultures, an examination of stool for ova and parasites, and a test for detection of Clostridium difficile toxin if there had been a history of antibiotic use.
The approach should then vary depending on the clinical presentation. If the patient has mucousy, bloody diarrhea, suggesting colitis, a colonoscopy is the next step. If the patient has intermittent abdominal pain or growth failure without symptoms of colitis, a barium study of the upper gastrointestinal tract and distal small intestines should be performed. If there were no evidence of intestinal disease on the x-ray and I still had a clinical suspicion of IBD, I would probably perform endoscopy of both the upper and lower gastrointestinal tract to make a tissue diagnosis.
Dr. Listernick: What is the utility of the newer serologie tests that are being touted as screening tools for detecting IBD?
Dr. Emerick: The main antibody tests are anti-neutrophil cytoplasmic antibody (pANCA) and anu-Saccharomyces cerevisiae antibody (ASCA). It was found that a very high percentage of patients with ulcerative colitis test positive for pANCA, whereas a similarly high percentage of patients with Crohn's disease are positive for ASCA. However, this testing has not been performed on a large series of controls. We also know that the positive predictive value of a test drops dramatically when used in a population that has a low incidence of disease.
Dr Listernick: In other words, you wouldn't recommend use of this test as a screening tool in a group of children who have chronic nonspecific abdominal pain without any other features suggestive of IBD?
Dr. Emerick: I may not, but other physicians use it differently. Positive serology might lead them into doing more invasive testing.
Dr. Listernick: What was the likelihood of this patient having IBD with all the negative testing to date?
Dr. Emerick: If the colonoscopy extended to the cecum and terminal ileum, IBD would have been extremely unlikely.
Dr. Listernick: What can we say about her anemia?
Deborah BrOWn7 MD, pediatric hematologist: She has a hypochromic, microcytic anemia. My next step would be to look at the blood smear. Assuming the smear looks the same, the most likely diagnosis would be iron deficiency. As she is Tanner 2 and obviously does not have menorrhagia, the most common cause would be gastrointestinal losses.
Dr. Listernick: Could this be the anemia of chronic disease/ inflammation?
Dr. Brown: Yes. The anemia of chronic inflammation may be either microcytic or normocytic. A low serum ferritin or high total iron binding capacity (TIBC) would support iron deficiency; a low TIBC would be seen in chronic inflammation. The ferritin is an acute phase reactant and may be elevated in acute inflammation, providing false reassurance about the possibility of iron deficiency.
Dr. Listernick: What about the lack of response to iron?
Dr. Brown: This could be due to decreased absorption of iron related to her underlying disease or to continued gastrointestinal blood loss. There's no way to tell at the moment.
Dr. Listernick: How do you interpret her conversion to a positive purified protein derivative (PPD)?
Ben Katz, MD, pediatric infectious disease specialist: Tuberculin skin test converters are at risk throughout their lives for developing tuberculosis (TB). Children younger than one year of age have a lifetime risk of developing TB of more than 40%. This risk decreases throughout life. An adolescent may have a 10% to 15% lifetime risk, whereas a 35 year old has about a 3% risk.
Dr. Listernick: Could active TB account for her symptoms?
Dr. Katz: Because she's had a negative chest x-ray, she would have to have had extrapulmonary TB. Abdominal tuberculosis in mesenteric lymph nodes is a rarely described phenomenon. However, she's had a very thorough abdominal evaluation. Certainly mesenteric lymphadenitis should have shown up on the CT scan.
Although 30% of TB cases in children may be extrapulmonary, the obvious sites, such as me bones and abdomen, have been evaluated. She certainly didn't have TB meningitis with the length of her symptoms and the lack of progression. Given all of the above, I don't believe she has active TB.
Stanford Shulman, MD, pediatrie infectious disease specialist: TB may involve any organ, as documented in previous decades. Since she was having back pain, it would have been very important to rule out spine disease. Although menstruating women may have pelvic osteomyelitis, she is prepubertal.
Dr. Li stern i ck: What about miliary TB?
Richard Shore, MD, pediatrie radiologist: Although miliary TB may be missed on a chest x-ray, it is readily identifiable by CT scan. TB of the spine should have been identified easily on either the bone scan or MRI of the spine.
Dr. Listernick: As back pain was a prominent symptom, I considered the possibility of ankylosing spondylitis. A test for the presence of HLA-B27 was negative.
Marissa Klein-Gittelman, pediatrie rheumatologist: Ifs a very reasonable thought. Males are at higher risk for the development of ankylosing spondylitis and the females who develop it tend to have less severe disease. There are many spondyloarthropathy in which individuals are HLA-B27 negative. However, 98% of white individuals with ankylosing spondylitis are HLA-B27 positive.
Dr. Listernick: We're at the point when she was first evaluated here. Her CBC and ESR were unchanged. All of her previous x-rays were reviewed by our radiologists and felt to be normal. How should we have proceeded at this point?
Dr. Katz: There's definitely a role for watchful waiting at times. Rather than jumping in and doing dozens of tests that aren't fruitful, repeated questioning and physical examinations may be more likely to yield useful information.
Robert Garofalo, MD, adolescent medicine physician: She definitely should have been tested for HIV, particularly as she has a positive PPD.
Dr. Listernick: I agree, although I didn't think of it at the time. For whatever reason, I thought, "This girl must have a chronic inflammatory /infectious process." I decided to repeat the gallium scan.
Dr. Shore: This child was scanned at 24 and 48 hours following the injection of gallium. At both times, there was persistently increased activity in the right lower quadrant, somewhat less intense at 48 hours. Gallium activity in the intestines is always difficult to interpret; gallium is normally excreted into the bowel. We often give laxatives to try to prove that the localization is not just in fecal matter. There was also increased activity in the sternum. The exact meaning of these findings was unclear.
Dr. Shulman: What is the utility of performing a gallium scan in these circumstances, when you have no idea what is going on but you suspect inflammation? I am particularly concerned because gallium scans deliver a significantly higher dose of radiation compared to other nuclear medicine tests.
Dr. Shore: Generally, gallium scintigraphy is unlikely to be helpful if the scan is performed to search for inflammation in children who have no localizing findings. However, if clinical suspicion is strong that an occult focus of inflammation exists, gallium scintigraphy is useful because it provides an overview of the entire body.
Dr. Listernick: I was still focused on the combination of abdominal pain, elevated sedimentation rate, refractory iron deficiency anemia, positive PPD, and the abnormal gallium scan. Even though the abdominal CT scan and colonoscopy were both normal, the possibility of an inflammatory process in the peritoneum seemed real. At my suggestion, the surgeons agreed to perform laparoscopy. Unfortunately, it was entirely normal.
Dr. Katz: I told you to leave her alone! Of course, in the era prior to noninvasive testing, laparotomy was a common modality used in establishing the etiology of fever of unknown origin.
Dr. Listernick: What should we have done now? It was approximately 7 weeks from the onset of her symptoms, which have not improved.
Dr. Katz: I would have treated her with isoniazid as a tuberculin skin test converter and followed her closely without any more testing unless there were changes in her condition. There is a real danger to blindly repeating tests and invasive procedures.
Dr. Shulman: Given her symptom complex, treating her with a single anti-TB drug would certainly have been a great way to induce isonicotine hydrazine (INH) resistance if she had active TB.
Dr, Brown: I would have continued checking the stools for blood loss while following her.
A. Todd Davis, MD, general academic pediatrician: When physicians are in this situation, if s helpful to explicitly point out to the parents that the human brain does not tolerate uncertainty or ambiguity very well.
Dr. Listernick: Despite all that's been said, I didn't feel that "watchful waiting" was the correct approach in this situation. I felt certain that there was an occult inflammatory process. Because of her back pain, I decided to repeat the MRI of her spine that had been performed 7 weeks earlier.
Dr. Shore: Her spine was normal. However, there was a serendipitous finding. The entire thoracic descending aorta is markedly dilated and innamed. I suspect that the abnormal activity in the area of the sternum on the gallium scan actually represented increased uptake of gallium by the aorta.
Or. Klein-Gitelman: This girl has Takayasu's arteritis. Angiography confirmed the aortitis as well as demonstrating left carotid artery stenosis. When I saw her, she definitely had a diminished left radial artery pulse.
Dr. Listernick: I admit that I didn't look carefully at die symmetry of her radial pulses.
Dr. Klein-Gitelman: Although nonspecific, her symptoms were classic for Takayasu's arteritis. The physician must have a high index of suspicion, as mere is no specific laboratory finding. An ophthalmologic examination is helpful at times, demonstrating retinal arterial narrowing, large arteriovenous anastamoses, or peripheral microaneurysms. Her eye examination was normal. Careful attention needs to be paid to the cardiovascular examination and the peripheral pulses.
This is an inflammatory disease of the large vessels, affecting the elastic fibers in the adventitia. Complications include aneurysmal dilatation as well as stenosis. These can lead to severe hypertension, heart failure, or stroke depending upon which vessels are involved. The mainstay of treatment is corticosteroids.
Dr. Listernick: What has been her clinical course?
Dr. Klein-Gitelman: She has had persistently active disease during the past few years that has been very difficult to control. In addition to the aorta and carotids, she has had involvement of the abdominal vasculature. She has required large doses of steroids because she did not respond to cyclophosphamide and became intolerant to the side effects she had with methotrexate.
She is currently receiving mycophenolate mofetil, an immune suppressant. If she does not respond, use of infliximab, a monoclonal antibody that binds to tumor necrosis factor alpha, will be considered. She has developed multiple compression fractures as a complication of the steroids.
Dr. Listernick: How do we account for the positive tuberculin test?
Dr. Shulman: Takayasu's arteritis was first reported in Japan, which had a high rate of TB in the population at the time. It has been reported to have an increased prevalence in other populations with high rates of TB, such as South Africa. It is quite unclear whether this represents a pathogenetic relationship or an epiphenomenon.
A second issue is the treatment of a tuberculin positive child who needs to receive substantial immunosuppression. Even though we had no evidence of active TB, it was decided that she should be treated with 3 anti-TB drugs for at least 6 months: isoniazid, rifampin, and ethambutol.
Dr. Klein-Gitelman: She still receives isoniazid, 5 years later. We monitor her pulmonary and central aortic pressures echocardiograms and periodic angiography, The caliber of her abnormal vessels is monitored annually with MR angiography.
Dr. Shulman: As far as establishing the diagnosis, if s good to be smart, but better to be lucky.
Dr. Listernick: No argument.