Special Cure
http://www.100md.com
《新英格兰医药杂志》
In this Journal feature, information about a real patient is presented in stages (boldface type) to an expert clinician, who responds to the information, sharing his or her reasoning with the reader (regular type). The authors' commentary follows.
A 42-year-old white man presented with a history of eight months of pain in his low back, hips, ankles, and feet. He had begun experiencing progressively severe pain during a 16-month period of incarceration. Therapy with nonsteroidal antiinflammatory medication had provided no significant improvement. The pain had become so severe that the patient had difficulty walking. He had otherwise felt well but had lost weight (from 76.2 to 67.6 kg) without a change in diet. He reported having had no fever, night sweats, diarrhea, anorexia, vomiting, abdominal distention, or dysphagia.
A number of characteristics distinguish this man's condition from that of the typical case of idiopathic low-back pain: his difficulty walking, the long duration of the condition despite conservative therapy, his weight loss, and the presence of severe pain throughout both lower extremities. These characteristics raise my concern about a neoplastic, infectious, or inflammatory disorder. His residence in prison poses certain health threats. Tuberculosis has a predilection for the thoracic and lumbar spine and can cause paralysis through vertebral destruction and compression of the spinal cord. Although tuberculosis could be manifested as an indolent monoarthritis, a symmetric bilateral polyarthritis would be decidedly uncommon. Trauma and fractures in the spine or lower extremities from altercations must also be ruled out. Another possibility is that he has an inflammatory disorder with polyarthritis. The list of possible disorders is quite long, but one of the first that comes to mind here is a seronegative spondylarthropathy, particularly reactive arthritis, which has a predilection for the lower extremities. It can be either acute and self-limited or chronic and progressive. However, weight loss is not a common feature of this disorder.
The patient had been born with gastroschisis (a congenital externalization of the bowel), which had been corrected surgically. Four years before the current evaluation, iron-deficiency anemia, hepatitis C, and hypothyroidism were diagnosed synchronously, and he was treated with oral iron and thyroxine. He had a distant history of intravenous drug use and episodic alcohol use and had a 20-pack-year history of smoking. He reported that his mother had died of stomach cancer and had required a colostomy at the age of 27 years for an unknown condition.
I would not expect that a congenital malformation would lead to a mechanical or physiological disorder four decades after a successful repair. However, I wonder whether the genotype associated with gastroschisis could have predisposed this patient to other disorders of the gastrointestinal tract later in life. Although iron deficiency in a man is almost always explained by gastrointestinal blood loss (e.g., due to peptic ulcer disease or gastritis), it may infrequently result from failure to absorb iron due to a disorder such as celiac sprue. Hypothyroidism is usually autoimmune in origin and has been described as an extrahepatic manifestation of hepatitis C infection. The patient's history of intravenous drug use raises the possibility of infection with the human immunodeficiency virus.
The family history of stomach and colon problems and the patient's congenital malformation and iron-deficiency anemia all center around the gastrointestinal tract. Weight loss without a change in diet raises the possibility of malabsorption; other important causes of unintended weight loss with a preserved appetite are hyperthyroidism and diabetes mellitus. In trying to reconcile these possibilities with the patient's progressive symptoms in the back and lower extremities, I am still considering a seronegative spondylarthropathy, since it can occur in association with inflammatory bowel disease and other gastrointestinal conditions. One argument against this line of reasoning, however, is the absence of gastrointestinal symptoms. Another possibility is that the symptoms in his lower extremities are caused by hepatitis C–related cryoglobulinemia, with neuropathy and arthralgia. I remain uncertain of the cause of his pain at this point.
The patient's temperature was 37.1°C, his heart rate 70 beats per minute, his blood pressure 115/70 mm Hg, and his respiratory rate 20 breaths per minute. He was thin and pale but did not appear to be in distress. The examination of the heart and lungs showed no abnormalities. His abdomen had a midline scar; there was no organomegaly or tenderness. His stool was guaiac-positive. He had normal joint mobility without effusion or warmth. His muscle strength was normal without muscle tenderness. Although he had marked pain in moving any joint, the pain could not be localized to the joint itself. His reflexes were symmetric and normal throughout. He had mild clubbing of his fingers. The rest of the examination was normal.
The guaiac-positive stool points to a mucosal gastrointestinal disorder that explains the patient's iron-deficiency anemia, a finding that indicates a need for endoscopic examination. The patient does not have motor weakness; instead, his inability to walk seems to stem from pain with movement. The examination does not suggest an inflammatory arthritis, yet he cannot tolerate movement of the affected joints. Although intra-articular disease is still possible, his symptoms make me consider a pathologic process in the surrounding structures, perhaps an enthesopathy or periostitis. The fact that his fingers are clubbed would prompt me to obtain a chest radiograph in order to evaluate the possibility of lung cancer, which is very commonly associated with clubbing. Cyanotic heart disease is another major cause of clubbing and is worth mentioning, since he had another congenital malformation; however, the absence of cardiopulmonary symptoms or corrective surgery during the past four decades essentially rules out that diagnosis. Clubbing can be an extraintestinal manifestation of inflammatory bowel disease. A systemic disorder, hypertrophic osteoarthropathy may be manifested as both clubbing and periostosis, a sometimes painful condition affecting the long bones.
The hemoglobin level was 6.3 g per deciliter, the hematocrit 21.7 percent, the mean corpuscular volume 67.7 μm3, and the red-cell distribution width 20.9 percent. His white-cell count was 8300 per cubic millimeter, with a normal differential count, and the platelet count was 221,000 per cubic millimeter. The erythrocyte sedimentation rate was 70 mm per hour, and the absolute reticulocyte count was 2 percent. Other laboratory results were as follows: serum calcium, 6.9 mg per deciliter (1.7 mmol per liter); albumin, 3.2 g per deciliter; phosphorus, 1.8 mg per deciliter (0.6 mmol per liter); total protein, 6.3 g per deciliter; alanine aminotransferase, 83 U per liter; aspartate aminotransferase, 146 U per liter; alkaline phosphatase, 445 U per liter; and total bilirubin, 0.2 mg per deciliter (3.4 mmol per liter).
The patient's profound anemia is probably an important contributor to his sense of weakness. The guaiac-positive stool, the low mean corpuscular volume, and the low reticulocyte count are all consistent with iron deficiency. I wonder whether he is still taking his prescribed iron-replacement therapy. An elevated red-cell distribution width, which reflects a wide variance in the size of the red-cell population, is of limited discriminatory value. It may reflect a heterogeneous cell population due to two simultaneous effects on erythropoiesis — microcytosis from iron deficiency and macrocytosis from folate deficiency, for example — or it may reflect a nonhomogeneous population of cells due to a single disorder, such as iron deficiency.
The combination of low levels of phosphorus and calcium is characteristic of vitamin D deficiency, which is likely to be a consequence of malabsorption caused by the patient's mucosal disease. The elevated alkaline phosphatase level probably reflects the increased turnover of bone from secondary hyperparathyroidism. The aminotransferase levels can be explained by hepatitis C infection and alcoholic liver disease.
What is the nature of the patient's mucosal disease? It can be difficult to differentiate between inflammatory bowel disease, which I considered initially, and celiac sprue, another likely diagnosis. Both disorders may have extraintestinal manifestations and can cause malabsorption and iron deficiency. The family history (with colostomy), elevated sedimentation rate, and clubbing are more suggestive of inflammatory bowel disease, whereas vitamin D malabsorption and autoimmune thyroiditis are more typically associated with celiac sprue. Iron deficiency may occur in either condition, although the guaiac-positive stool is more characteristic of inflammatory bowel disease.
A plain radiograph of the patient's pelvis showed marked osteopenia (Figure 1). Dual-energy x-ray absorptiometry revealed a T score of –5.9, indicating that the patient's bone density was nearly 6 SD below the mean for normal young men. The results of laboratory tests were as follows: serum ferritin, 5 ng per milliliter (normal range, 22 to 322); free thyroxine, 0.83 ng per deciliter (10.68 pmol per liter; normal range, 0.84 to 1.51 ng per deciliter ); and thyrotropin, 16.4 μU per milliliter (normal range, 0.35 to 5.50). The levels of vitamin B12 and folate were normal; HLA-B27 was not detected.
Figure 1. Radiographs of the Pelvic Area.
A plain radiograph of the patient's pelvic area shows severe osteopenia (Panel A). A radiograph of the pelvic area in a person of the same age and sex as the patient shows normal bone mineralization (Panel B).
Bone densitometry reveals a severe decrease in bone density. This is probably the result of osteomalacia from vitamin D deficiency. The low level of serum ferritin confirms profound iron deficiency. The patient's hypothyroidism persists, and although it is mild, it may also be contributing to his sense of weakness. Although HLA-B27 is associated with axial arthritis (spondylitis–sacroiliitis) rather than the peripheral arthritis of spondylarthropathy, this test has limited diagnostic utility. My suspicion that the diagnosis is celiac sprue has increased substantially.
The parathyroid hormone level was extremely elevated, at 613 pg per milliliter (normal range, 15 to 65), and the 25-hydroxyvitamin D level was undetectable. The total cholesterol level was 65 mg per deciliter (1.68 mmol per liter). Endoscopy revealed a nonbleeding duodenal ulcer and a marked "scalloping" of the duodenal folds.
These values confirm hypovitaminosis D and associated secondary hyperparathyroidism. Marked hypolipidemia is seen in advanced cancer with cachexia or severe malnutrition, but in this case it is explained by malabsorption. The ulcer, which I did not expect, may be explained by infection with Helicobacter pylori or the patient's use of nonsteroidal antiinflammatory drugs, but other causes of ulceration also merit consideration. With inflammatory bowel disease in the differential diagnosis, a biopsy would be important to rule out noncaseating granulomas, which are characteristic of Crohn's disease. Given the lymphocytic infiltration of the mucosa in celiac sprue, and the associated increase in the risk of intestinal cancer, my primary concern would be to rule out cancer, particularly lymphoma, as the cause of this ulcer.
A biopsy of the ulcer and the nonulcerated mucosa and tests for antiendomysial or antigliadin antibodies will be the definitive studies. A diffuse mucosal disease of the small intestine in the setting of severe malabsorption (with deficiencies in iron, vitamin D, and cholesterol) points to celiac sprue. Crohn's disease, which usually affects the distal small intestine, is less frequently associated with significant malabsorption. Less common enteropathies — such as tropical sprue, bacterial overgrowth, a diffuse gastrointestinal lymphoma, or Whipple's disease (even with its predilection for men and its associated arthralgia) — are unlikely.
Biopsy of the small bowel revealed marked villous atrophy (Figure 2). A test for antigliadin IgA antibodies was positive, at 209.6 enzyme-linked immunosorbent assay units, and a test for antiendomysial IgA antibodies was positive at a titer of 1:160. A diagnosis of celiac sprue was made. The patient was given a transfusion of packed red cells and intravenous iron, calcium, and vitamin D. He was discharged back to prison on a gluten-free diet.
Figure 2. Biopsy Specimens of the Duodenum.
The duodenal-biopsy specimen from the patient (Panel A) reveals degeneration of surface epithelial cells with an increased number of intraepithelial lymphocytes, total villous atrophy, crypt hyperplasia, and chronic inflammation. Panel B shows normal duodenal mucosa with tall villi, normal lamina propria, and shallow crypts.
After the patient had been on this diet (which was prepared specifically for him by the prison cafeteria staff) for four months, his pain had resolved and he had regained significant weight. His hemoglobin and calcium levels had normalized.
Commentary
With a prevalence of almost 1 in 100 in the general population, celiac sprue is surprisingly common.1,2,3 Unfortunately, it is a disease that also often leads to diagnostic confusion, particularly if patients do not present with chronic diarrhea and steatorrhea. This case illustrates both why the diagnosis of sprue may be challenging and how a constellation of seemingly unrelated findings led to the recognition of a pattern consistent with celiac disease.
The presenting symptom of progressive hip and leg pain with difficulty walking led the discussant to consider causes of both localized and diffuse bone pain. He initially suspected that the patient had inflammatory bowel disease complicated by a seronegative spondylarthropathy but also considered malabsorption, given the patient's history of weight loss and iron-deficiency anemia. Although both inflammatory bowel disease and celiac sprue were possibilities, the discussant settled on sprue after the discovery of vitamin D deficiency and bone loss (with osteomalacia the presumed cause of his bone pain).
A lack of awareness of the diverse clinical manifestations of celiac sprue may have led many clinicians to miss the diagnosis. In this case, the patient did not report abdominal pain or diarrhea (symptoms widely believed to be essential to the diagnosis), and the stool was guaiac-positive. However, abdominal pain or diarrhea each occurs in only about one third of patients with celiac sprue,2 and occult blood is present in the stool in almost half of all patients with the disease.4 In the United States, iron-deficiency anemia is the most common presentation of sprue.5 Other common symptoms include fatigue, joint pain, diarrhea, constipation, and abdominal pain. Coexisting autoimmune diseases are common, such as hypothyroidism (in 10 percent of patients with celiac sprue), hyperthyroidism (4 percent), and type 1 diabetes mellitus (7 percent).2,6 Gastrointestinal neoplasms and T-cell lymphomas occur in up to 11 percent of patients.7 In patients with hepatitis C, the risk of celiac disease is increased by a factor of three,8 but there is no recognized association between gastroschisis and sprue.
Although the diagnosis of sprue often poses challenges, the pathophysiology of the disease is well understood. Celiac sprue is a consequence of intestinal inflammation initiated in response to the wheat protein gluten. After proteolysis and activation by brush-border aminotransferases, the resultant polypeptides initiate activation of helper T cells, the release of cytokines, inflammation, and subsequent villous atrophy.9 The diagnosis is usually made after a duodenal biopsy reveals villous atrophy with a lymphocytic infiltration in a patient with characteristic antibodies. If seen on endoscopy, scalloping of intestinal mucosa has a specificity of more than 90 percent.10 The test for antitransglutaminase IgA antibodies has a sensitivity of 93 to 98 percent and a specificity of 94 to 100 percent5 and is considered a test of choice for initially evaluating suspected cases and for screening high-risk patients.5
This clinical problem-solving exercise exemplifies the importance of efficient pattern recognition in making a correct diagnosis. Clinicians rely on a back-and-forth process of data collection and interpretation that allows for a rapid narrowing of possibilities to a few preliminary hypotheses. But the hypotheses are only as good as the templates from which they are drawn. Astute clinicians usually use a process of probabilistic reasoning (often implicit), whereby clinical findings increase or decrease the likelihood of specific disease entities.11 The assignment of correct prior probabilities occurs only if the clinician has an accurate template of a given disease to compare with information as it is gathered. For example, most clinicians would correctly assign a very high probability to a diagnosis of hyperthyroidism in a young woman presenting with proptosis, hyperactivity, fine tremor, and diarrhea. These symptoms bring to mind the combination of signs and symptoms (i.e., the hyperthyroidism template) that is learned in medical training. In elderly patients, on the other hand, hyperthyroidism is often characterized by apathy, and this otherwise reliable template unravels, increasing the chance of missing the diagnosis.
In the present case, certain clues pointed toward the ultimate diagnosis of sprue. An expert clinician with a thorough knowledge of the signs, symptoms, associated conditions, and epidemiology of sprue might have suspected celiac disease after just the initial history — but only if armed with a correct template. The discussant's line of reasoning began with noting that a young man with iron-deficiency anemia almost certainly has a gastrointestinal lesion. In fact, up to 5 percent of cases of iron-deficiency anemia can be attributed to celiac disease.12,13,14 Add in weight loss, bony pain, and a history of hypothyroidism, and one could argue that celiac sprue becomes the most likely possibility. The high prevalence of the disease provides additional support for this argument.
Skilled clinicians have a finely tuned ability to take information as it is presented and identify a pattern pointing to the ultimate diagnosis. An astute reader could have suspected the diagnosis in this case after reading the title alone. Celiac disease does indeed have a special cure: dietary modification. As an anagram for "celiac sprue," "special cure" was a hidden pattern awaiting recognition.
Supported by a Career Development Award from the Health Services Research and Development Program of the Department of Veterans Affairs and a Patient Safety Developmental Center Grant (P20-HS11540) from the Agency for Healthcare Research and Quality (both to Dr. Saint).
Source Information
From the Department of Medicine, Legacy Emanuel and Legacy Good Samaritan Hospitals, Portland, Oreg. (R.J.H., D.J.G.); the Department of Medicine, San Francisco Veterans Affairs Medical Center and the University of California at San Francisco School of Medicine, San Francisco (G.D.); and the Ann Arbor Veterans Affairs Health Services Research and Development Center of Excellence; the Patient Safety Enhancement Program, University of Michigan Health System; and the Department of Internal Medicine, University of Michigan Medical School — all in Ann Arbor (S.S.).
Address reprint requests to Dr. Dhaliwal at the San Francisco Veterans Affairs Medical Center, 4150 Clement St. (111), San Francisco, CA 94121, or at gurpreet.dhaliwal@med.va.gov.
References
M?ki M, Mustalahti K, Kokkonen J, et al. Prevalence of celiac disease among children in Finland. N Engl J Med 2003;348:2517-2524.
Fasano A, Berti I, Gerarduzzi T, et al. Prevalence of celiac disease in at-risk and not-at-risk groups in the United States: a large multicenter study. Arch Intern Med 2003;163:286-292.
Hin H, Bird G, Fisher P, Mahy N, Jewell D. Coeliac disease in primary care: case finding study. BMJ 1999;318:164-167.
Fine KD. The prevalence of occult gastrointestinal bleeding in celiac sprue. N Engl J Med 1996;334:1163-1167.
Farrell RJ, Kelly CP. Celiac sprue. N Engl J Med 2002;346:180-188.
Counsell CE, Taha A, Ruddell WS. Coeliac disease and autoimmune thyroid disease. Gut 1994;35:844-846.
Green PH, Fleischauer AT, Bhagat G, Goyal R, Jabri B, Neugut AI. Risk of malignancy in patients with celiac disease. Am J Med 2003;115:191-195.
Fine KD, Ogunji F, Saloum Y, Beharry S, Crippin J, Weinstein J. Celiac sprue: another autoimmune syndrome associated with hepatitis C. Am J Gastroenterol 2001;96:138-145.
McManus R, Kelleher D. Celiac disease -- the villain unmasked? N Engl J Med 2003;348:2573-2574.
Oxenteko AS, Grisolano SW, Murray JA, Burgart LJ, Dierkhising RA, Alexander JA. The insensitivity of endoscopic markers in celiac disease. Am J Gastroenterol 2002;97:933-938.
Kassirer JP. Diagnostic reasoning. Ann Intern Med 1989;110:893-900.
Karnam US, Felder LR, Raskin JB. Prevalence of occult celiac disease in patients with iron-deficiency anemia: a prospective study. South Med J 2004;97:30-34.
McIntyre AS, Long RG. Prospective survey of investigations in outpatients referred with iron deficiency anaemia. Gut 1993;34:1102-1107.
Corazza GR, Valentini RA, Andreani ML, et al. Subclinical coeliac disease is a frequent cause of iron-deficiency anaemia. Scand J Gastroenterol 1995;30:153-156.(Robert J. Hoffman, M.D., )
A 42-year-old white man presented with a history of eight months of pain in his low back, hips, ankles, and feet. He had begun experiencing progressively severe pain during a 16-month period of incarceration. Therapy with nonsteroidal antiinflammatory medication had provided no significant improvement. The pain had become so severe that the patient had difficulty walking. He had otherwise felt well but had lost weight (from 76.2 to 67.6 kg) without a change in diet. He reported having had no fever, night sweats, diarrhea, anorexia, vomiting, abdominal distention, or dysphagia.
A number of characteristics distinguish this man's condition from that of the typical case of idiopathic low-back pain: his difficulty walking, the long duration of the condition despite conservative therapy, his weight loss, and the presence of severe pain throughout both lower extremities. These characteristics raise my concern about a neoplastic, infectious, or inflammatory disorder. His residence in prison poses certain health threats. Tuberculosis has a predilection for the thoracic and lumbar spine and can cause paralysis through vertebral destruction and compression of the spinal cord. Although tuberculosis could be manifested as an indolent monoarthritis, a symmetric bilateral polyarthritis would be decidedly uncommon. Trauma and fractures in the spine or lower extremities from altercations must also be ruled out. Another possibility is that he has an inflammatory disorder with polyarthritis. The list of possible disorders is quite long, but one of the first that comes to mind here is a seronegative spondylarthropathy, particularly reactive arthritis, which has a predilection for the lower extremities. It can be either acute and self-limited or chronic and progressive. However, weight loss is not a common feature of this disorder.
The patient had been born with gastroschisis (a congenital externalization of the bowel), which had been corrected surgically. Four years before the current evaluation, iron-deficiency anemia, hepatitis C, and hypothyroidism were diagnosed synchronously, and he was treated with oral iron and thyroxine. He had a distant history of intravenous drug use and episodic alcohol use and had a 20-pack-year history of smoking. He reported that his mother had died of stomach cancer and had required a colostomy at the age of 27 years for an unknown condition.
I would not expect that a congenital malformation would lead to a mechanical or physiological disorder four decades after a successful repair. However, I wonder whether the genotype associated with gastroschisis could have predisposed this patient to other disorders of the gastrointestinal tract later in life. Although iron deficiency in a man is almost always explained by gastrointestinal blood loss (e.g., due to peptic ulcer disease or gastritis), it may infrequently result from failure to absorb iron due to a disorder such as celiac sprue. Hypothyroidism is usually autoimmune in origin and has been described as an extrahepatic manifestation of hepatitis C infection. The patient's history of intravenous drug use raises the possibility of infection with the human immunodeficiency virus.
The family history of stomach and colon problems and the patient's congenital malformation and iron-deficiency anemia all center around the gastrointestinal tract. Weight loss without a change in diet raises the possibility of malabsorption; other important causes of unintended weight loss with a preserved appetite are hyperthyroidism and diabetes mellitus. In trying to reconcile these possibilities with the patient's progressive symptoms in the back and lower extremities, I am still considering a seronegative spondylarthropathy, since it can occur in association with inflammatory bowel disease and other gastrointestinal conditions. One argument against this line of reasoning, however, is the absence of gastrointestinal symptoms. Another possibility is that the symptoms in his lower extremities are caused by hepatitis C–related cryoglobulinemia, with neuropathy and arthralgia. I remain uncertain of the cause of his pain at this point.
The patient's temperature was 37.1°C, his heart rate 70 beats per minute, his blood pressure 115/70 mm Hg, and his respiratory rate 20 breaths per minute. He was thin and pale but did not appear to be in distress. The examination of the heart and lungs showed no abnormalities. His abdomen had a midline scar; there was no organomegaly or tenderness. His stool was guaiac-positive. He had normal joint mobility without effusion or warmth. His muscle strength was normal without muscle tenderness. Although he had marked pain in moving any joint, the pain could not be localized to the joint itself. His reflexes were symmetric and normal throughout. He had mild clubbing of his fingers. The rest of the examination was normal.
The guaiac-positive stool points to a mucosal gastrointestinal disorder that explains the patient's iron-deficiency anemia, a finding that indicates a need for endoscopic examination. The patient does not have motor weakness; instead, his inability to walk seems to stem from pain with movement. The examination does not suggest an inflammatory arthritis, yet he cannot tolerate movement of the affected joints. Although intra-articular disease is still possible, his symptoms make me consider a pathologic process in the surrounding structures, perhaps an enthesopathy or periostitis. The fact that his fingers are clubbed would prompt me to obtain a chest radiograph in order to evaluate the possibility of lung cancer, which is very commonly associated with clubbing. Cyanotic heart disease is another major cause of clubbing and is worth mentioning, since he had another congenital malformation; however, the absence of cardiopulmonary symptoms or corrective surgery during the past four decades essentially rules out that diagnosis. Clubbing can be an extraintestinal manifestation of inflammatory bowel disease. A systemic disorder, hypertrophic osteoarthropathy may be manifested as both clubbing and periostosis, a sometimes painful condition affecting the long bones.
The hemoglobin level was 6.3 g per deciliter, the hematocrit 21.7 percent, the mean corpuscular volume 67.7 μm3, and the red-cell distribution width 20.9 percent. His white-cell count was 8300 per cubic millimeter, with a normal differential count, and the platelet count was 221,000 per cubic millimeter. The erythrocyte sedimentation rate was 70 mm per hour, and the absolute reticulocyte count was 2 percent. Other laboratory results were as follows: serum calcium, 6.9 mg per deciliter (1.7 mmol per liter); albumin, 3.2 g per deciliter; phosphorus, 1.8 mg per deciliter (0.6 mmol per liter); total protein, 6.3 g per deciliter; alanine aminotransferase, 83 U per liter; aspartate aminotransferase, 146 U per liter; alkaline phosphatase, 445 U per liter; and total bilirubin, 0.2 mg per deciliter (3.4 mmol per liter).
The patient's profound anemia is probably an important contributor to his sense of weakness. The guaiac-positive stool, the low mean corpuscular volume, and the low reticulocyte count are all consistent with iron deficiency. I wonder whether he is still taking his prescribed iron-replacement therapy. An elevated red-cell distribution width, which reflects a wide variance in the size of the red-cell population, is of limited discriminatory value. It may reflect a heterogeneous cell population due to two simultaneous effects on erythropoiesis — microcytosis from iron deficiency and macrocytosis from folate deficiency, for example — or it may reflect a nonhomogeneous population of cells due to a single disorder, such as iron deficiency.
The combination of low levels of phosphorus and calcium is characteristic of vitamin D deficiency, which is likely to be a consequence of malabsorption caused by the patient's mucosal disease. The elevated alkaline phosphatase level probably reflects the increased turnover of bone from secondary hyperparathyroidism. The aminotransferase levels can be explained by hepatitis C infection and alcoholic liver disease.
What is the nature of the patient's mucosal disease? It can be difficult to differentiate between inflammatory bowel disease, which I considered initially, and celiac sprue, another likely diagnosis. Both disorders may have extraintestinal manifestations and can cause malabsorption and iron deficiency. The family history (with colostomy), elevated sedimentation rate, and clubbing are more suggestive of inflammatory bowel disease, whereas vitamin D malabsorption and autoimmune thyroiditis are more typically associated with celiac sprue. Iron deficiency may occur in either condition, although the guaiac-positive stool is more characteristic of inflammatory bowel disease.
A plain radiograph of the patient's pelvis showed marked osteopenia (Figure 1). Dual-energy x-ray absorptiometry revealed a T score of –5.9, indicating that the patient's bone density was nearly 6 SD below the mean for normal young men. The results of laboratory tests were as follows: serum ferritin, 5 ng per milliliter (normal range, 22 to 322); free thyroxine, 0.83 ng per deciliter (10.68 pmol per liter; normal range, 0.84 to 1.51 ng per deciliter ); and thyrotropin, 16.4 μU per milliliter (normal range, 0.35 to 5.50). The levels of vitamin B12 and folate were normal; HLA-B27 was not detected.
Figure 1. Radiographs of the Pelvic Area.
A plain radiograph of the patient's pelvic area shows severe osteopenia (Panel A). A radiograph of the pelvic area in a person of the same age and sex as the patient shows normal bone mineralization (Panel B).
Bone densitometry reveals a severe decrease in bone density. This is probably the result of osteomalacia from vitamin D deficiency. The low level of serum ferritin confirms profound iron deficiency. The patient's hypothyroidism persists, and although it is mild, it may also be contributing to his sense of weakness. Although HLA-B27 is associated with axial arthritis (spondylitis–sacroiliitis) rather than the peripheral arthritis of spondylarthropathy, this test has limited diagnostic utility. My suspicion that the diagnosis is celiac sprue has increased substantially.
The parathyroid hormone level was extremely elevated, at 613 pg per milliliter (normal range, 15 to 65), and the 25-hydroxyvitamin D level was undetectable. The total cholesterol level was 65 mg per deciliter (1.68 mmol per liter). Endoscopy revealed a nonbleeding duodenal ulcer and a marked "scalloping" of the duodenal folds.
These values confirm hypovitaminosis D and associated secondary hyperparathyroidism. Marked hypolipidemia is seen in advanced cancer with cachexia or severe malnutrition, but in this case it is explained by malabsorption. The ulcer, which I did not expect, may be explained by infection with Helicobacter pylori or the patient's use of nonsteroidal antiinflammatory drugs, but other causes of ulceration also merit consideration. With inflammatory bowel disease in the differential diagnosis, a biopsy would be important to rule out noncaseating granulomas, which are characteristic of Crohn's disease. Given the lymphocytic infiltration of the mucosa in celiac sprue, and the associated increase in the risk of intestinal cancer, my primary concern would be to rule out cancer, particularly lymphoma, as the cause of this ulcer.
A biopsy of the ulcer and the nonulcerated mucosa and tests for antiendomysial or antigliadin antibodies will be the definitive studies. A diffuse mucosal disease of the small intestine in the setting of severe malabsorption (with deficiencies in iron, vitamin D, and cholesterol) points to celiac sprue. Crohn's disease, which usually affects the distal small intestine, is less frequently associated with significant malabsorption. Less common enteropathies — such as tropical sprue, bacterial overgrowth, a diffuse gastrointestinal lymphoma, or Whipple's disease (even with its predilection for men and its associated arthralgia) — are unlikely.
Biopsy of the small bowel revealed marked villous atrophy (Figure 2). A test for antigliadin IgA antibodies was positive, at 209.6 enzyme-linked immunosorbent assay units, and a test for antiendomysial IgA antibodies was positive at a titer of 1:160. A diagnosis of celiac sprue was made. The patient was given a transfusion of packed red cells and intravenous iron, calcium, and vitamin D. He was discharged back to prison on a gluten-free diet.
Figure 2. Biopsy Specimens of the Duodenum.
The duodenal-biopsy specimen from the patient (Panel A) reveals degeneration of surface epithelial cells with an increased number of intraepithelial lymphocytes, total villous atrophy, crypt hyperplasia, and chronic inflammation. Panel B shows normal duodenal mucosa with tall villi, normal lamina propria, and shallow crypts.
After the patient had been on this diet (which was prepared specifically for him by the prison cafeteria staff) for four months, his pain had resolved and he had regained significant weight. His hemoglobin and calcium levels had normalized.
Commentary
With a prevalence of almost 1 in 100 in the general population, celiac sprue is surprisingly common.1,2,3 Unfortunately, it is a disease that also often leads to diagnostic confusion, particularly if patients do not present with chronic diarrhea and steatorrhea. This case illustrates both why the diagnosis of sprue may be challenging and how a constellation of seemingly unrelated findings led to the recognition of a pattern consistent with celiac disease.
The presenting symptom of progressive hip and leg pain with difficulty walking led the discussant to consider causes of both localized and diffuse bone pain. He initially suspected that the patient had inflammatory bowel disease complicated by a seronegative spondylarthropathy but also considered malabsorption, given the patient's history of weight loss and iron-deficiency anemia. Although both inflammatory bowel disease and celiac sprue were possibilities, the discussant settled on sprue after the discovery of vitamin D deficiency and bone loss (with osteomalacia the presumed cause of his bone pain).
A lack of awareness of the diverse clinical manifestations of celiac sprue may have led many clinicians to miss the diagnosis. In this case, the patient did not report abdominal pain or diarrhea (symptoms widely believed to be essential to the diagnosis), and the stool was guaiac-positive. However, abdominal pain or diarrhea each occurs in only about one third of patients with celiac sprue,2 and occult blood is present in the stool in almost half of all patients with the disease.4 In the United States, iron-deficiency anemia is the most common presentation of sprue.5 Other common symptoms include fatigue, joint pain, diarrhea, constipation, and abdominal pain. Coexisting autoimmune diseases are common, such as hypothyroidism (in 10 percent of patients with celiac sprue), hyperthyroidism (4 percent), and type 1 diabetes mellitus (7 percent).2,6 Gastrointestinal neoplasms and T-cell lymphomas occur in up to 11 percent of patients.7 In patients with hepatitis C, the risk of celiac disease is increased by a factor of three,8 but there is no recognized association between gastroschisis and sprue.
Although the diagnosis of sprue often poses challenges, the pathophysiology of the disease is well understood. Celiac sprue is a consequence of intestinal inflammation initiated in response to the wheat protein gluten. After proteolysis and activation by brush-border aminotransferases, the resultant polypeptides initiate activation of helper T cells, the release of cytokines, inflammation, and subsequent villous atrophy.9 The diagnosis is usually made after a duodenal biopsy reveals villous atrophy with a lymphocytic infiltration in a patient with characteristic antibodies. If seen on endoscopy, scalloping of intestinal mucosa has a specificity of more than 90 percent.10 The test for antitransglutaminase IgA antibodies has a sensitivity of 93 to 98 percent and a specificity of 94 to 100 percent5 and is considered a test of choice for initially evaluating suspected cases and for screening high-risk patients.5
This clinical problem-solving exercise exemplifies the importance of efficient pattern recognition in making a correct diagnosis. Clinicians rely on a back-and-forth process of data collection and interpretation that allows for a rapid narrowing of possibilities to a few preliminary hypotheses. But the hypotheses are only as good as the templates from which they are drawn. Astute clinicians usually use a process of probabilistic reasoning (often implicit), whereby clinical findings increase or decrease the likelihood of specific disease entities.11 The assignment of correct prior probabilities occurs only if the clinician has an accurate template of a given disease to compare with information as it is gathered. For example, most clinicians would correctly assign a very high probability to a diagnosis of hyperthyroidism in a young woman presenting with proptosis, hyperactivity, fine tremor, and diarrhea. These symptoms bring to mind the combination of signs and symptoms (i.e., the hyperthyroidism template) that is learned in medical training. In elderly patients, on the other hand, hyperthyroidism is often characterized by apathy, and this otherwise reliable template unravels, increasing the chance of missing the diagnosis.
In the present case, certain clues pointed toward the ultimate diagnosis of sprue. An expert clinician with a thorough knowledge of the signs, symptoms, associated conditions, and epidemiology of sprue might have suspected celiac disease after just the initial history — but only if armed with a correct template. The discussant's line of reasoning began with noting that a young man with iron-deficiency anemia almost certainly has a gastrointestinal lesion. In fact, up to 5 percent of cases of iron-deficiency anemia can be attributed to celiac disease.12,13,14 Add in weight loss, bony pain, and a history of hypothyroidism, and one could argue that celiac sprue becomes the most likely possibility. The high prevalence of the disease provides additional support for this argument.
Skilled clinicians have a finely tuned ability to take information as it is presented and identify a pattern pointing to the ultimate diagnosis. An astute reader could have suspected the diagnosis in this case after reading the title alone. Celiac disease does indeed have a special cure: dietary modification. As an anagram for "celiac sprue," "special cure" was a hidden pattern awaiting recognition.
Supported by a Career Development Award from the Health Services Research and Development Program of the Department of Veterans Affairs and a Patient Safety Developmental Center Grant (P20-HS11540) from the Agency for Healthcare Research and Quality (both to Dr. Saint).
Source Information
From the Department of Medicine, Legacy Emanuel and Legacy Good Samaritan Hospitals, Portland, Oreg. (R.J.H., D.J.G.); the Department of Medicine, San Francisco Veterans Affairs Medical Center and the University of California at San Francisco School of Medicine, San Francisco (G.D.); and the Ann Arbor Veterans Affairs Health Services Research and Development Center of Excellence; the Patient Safety Enhancement Program, University of Michigan Health System; and the Department of Internal Medicine, University of Michigan Medical School — all in Ann Arbor (S.S.).
Address reprint requests to Dr. Dhaliwal at the San Francisco Veterans Affairs Medical Center, 4150 Clement St. (111), San Francisco, CA 94121, or at gurpreet.dhaliwal@med.va.gov.
References
M?ki M, Mustalahti K, Kokkonen J, et al. Prevalence of celiac disease among children in Finland. N Engl J Med 2003;348:2517-2524.
Fasano A, Berti I, Gerarduzzi T, et al. Prevalence of celiac disease in at-risk and not-at-risk groups in the United States: a large multicenter study. Arch Intern Med 2003;163:286-292.
Hin H, Bird G, Fisher P, Mahy N, Jewell D. Coeliac disease in primary care: case finding study. BMJ 1999;318:164-167.
Fine KD. The prevalence of occult gastrointestinal bleeding in celiac sprue. N Engl J Med 1996;334:1163-1167.
Farrell RJ, Kelly CP. Celiac sprue. N Engl J Med 2002;346:180-188.
Counsell CE, Taha A, Ruddell WS. Coeliac disease and autoimmune thyroid disease. Gut 1994;35:844-846.
Green PH, Fleischauer AT, Bhagat G, Goyal R, Jabri B, Neugut AI. Risk of malignancy in patients with celiac disease. Am J Med 2003;115:191-195.
Fine KD, Ogunji F, Saloum Y, Beharry S, Crippin J, Weinstein J. Celiac sprue: another autoimmune syndrome associated with hepatitis C. Am J Gastroenterol 2001;96:138-145.
McManus R, Kelleher D. Celiac disease -- the villain unmasked? N Engl J Med 2003;348:2573-2574.
Oxenteko AS, Grisolano SW, Murray JA, Burgart LJ, Dierkhising RA, Alexander JA. The insensitivity of endoscopic markers in celiac disease. Am J Gastroenterol 2002;97:933-938.
Kassirer JP. Diagnostic reasoning. Ann Intern Med 1989;110:893-900.
Karnam US, Felder LR, Raskin JB. Prevalence of occult celiac disease in patients with iron-deficiency anemia: a prospective study. South Med J 2004;97:30-34.
McIntyre AS, Long RG. Prospective survey of investigations in outpatients referred with iron deficiency anaemia. Gut 1993;34:1102-1107.
Corazza GR, Valentini RA, Andreani ML, et al. Subclinical coeliac disease is a frequent cause of iron-deficiency anaemia. Scand J Gastroenterol 1995;30:153-156.(Robert J. Hoffman, M.D., )