PRESENTATION OF CASE
Dr. Franklin W. Huang (Medicine): A 24-year-old woman was transferred to this hospital because of abdominal pain, vomiting, ascites, and shock.
The patient had been well until 9 days before admission, when fatigue, fevers, headache, and diffuse body aches developed, followed by a productive cough and vague pain in the abdomen and lower back. Three days before admission, nausea and vomiting developed, and oral intake decreased. Late in the evening of the next day, she was admitted to another hospital.
On evaluation, she appeared tired and restless; the blood pressure was 98/58 mm Hg, the pulse 122 beats per minute, the temperature 36.7°C, and the respiratory rate 22 breaths per minute. The abdomen was distended and diffusely tender, with normal bowel sounds; the remainder of the examination was normal. A rapid test for influenza was negative; other laboratory-test results are shown in Table 1TABLE 1
Laboratory Data.
. Intravenous crystalloid fluids were administered. During the next 3 hours, her abdominal pain increased (she rated it as 8 on a scale of 1 to 10, where 10 is the most severe), with increased abdominal distention and tenderness; ketorolac tromethamine was administered. A nasogastric tube was inserted, after which the hypotension worsened. She was admitted to the intensive care unit at the other hospital. Specimens of blood were sent for culture, and a central venous catheter was inserted; intravenous fluids and ertapenem were administered. Results on transthoracic echocardiography were normal. Computed tomography (CT) of the abdomen and pelvis after the administration of intravenous contrast material showed a large amount of ascites, no bowel dilatation, and concentric wall thickening of a long segment of the distal small bowel. Numerous slightly enlarged, avidly enhancing lymph nodes were present at the root of the small-bowel mesentery and in the retroperitoneum. Exploratory laparotomy revealed ascites (3 liters), which was drained; nodular, hemorrhagic lesions in the mesentery; and two areas of necrotic small bowel. Biopsy specimens were obtained from the mesenteric lesions, and the necrotic bowel and appendix were resected. Gram-positive rods, thought to be contaminants, grew after less than 15 hours of incubation from two of two blood-culture bottles containing blood obtained on admission. Cultures of the ascites remained sterile.
On the second day, hypotension and tachycardia persisted; hydroxyethyl starch was administered, and pressors were begun. The patient was transferred to this hospital.
The patient reported feeling tired and short of breath and having abdominal pain. Six months earlier, while working on an organic farm, an edematous nodule had developed on her left posterior calf, which she attributed to an insect bite. The lesion was raised, warm, and tender and reached 4 cm in diameter despite a 10-day course of amoxicillin–clavulanate. Incision and drainage produced a large amount of seropurulent fluid. Gram's staining revealed no organisms, and cultures were sterile. Ciprofloxacin was administered. The lesion reportedly drained gray fluid for several weeks; at follow-up 2 months before admission, it had resolved. The patient was taking an oral contraceptive. She had no allergies. She lived with her family, worked with children, and was active outdoors. She followed a vegan diet, drank alcohol infrequently, and did not smoke or use illicit drugs. She was not currently sexually active and reported no new foods, exposures, or travel beyond New England. There was a maternal family history of hypercoagulable disorders.
On examination, the patient was alert and appeared uncomfortable. The blood pressure was 129/70 mm Hg (while she was on pressors), the pulse 147 beats per minute, the body temperature 37.3°C, the respiratory rate 35 breaths per minute, and the oxygen saturation 99% while she was breathing supplemental oxygen. On auscultation of the lungs, breath sounds were diminished in intensity. The abdomen was distended and diffusely tender, with a clean midline surgical excision. There was 2+ peripheral edema and a scar (1.5 cm in diameter) on the left calf. The remainder of the examination was normal. Testing for antineutrophil cytoplasmic antibodies (ANCA) was negative; other test results are shown in Table 1. Urinalysis showed 2+ blood and trace protein, 20 to 50 red cells and 0 to 2 white cells per high-power field, and 20 to 100 hyaline casts per low-power field. A chest radiograph showed low lung volumes and layering of pleural effusions, and an abdominal radiograph showed no bowel obstruction or free air. An electrocardiogram revealed sinus rhythm at 156 beats per minute. She was admitted to the intensive care unit.
During the first day, pressors, intravenous fluids, ampicillin, ciprofloxacin, clindamycin, vancomycin, doxycycline, and peramivir were administered. Anemia, thrombocytopenia, and coagulopathy developed (Table 1), requiring transfusions of red cells, fresh-frozen plasma, platelets, and cryoglobulin. A chest radiograph obtained 8 hours after arrival showed large, layering effusions. A transthoracic echocardiogram showed a normal ejection fraction, no valvular vegetations, and findings consistent with right atrial volume overload and right ventricular systolic hypertension. Approximately 11 hours after the patient's arrival, the trachea was intubated because of worsening respiratory distress and abdominal distention. Refractory hypotension developed, requiring maximum vasopressor support. Paracentesis with removal of bloody ascites (3 liters) was performed, with lessening of the abdominal distention. Analysis of the orange, turbid fluid revealed 1425 white cells per cubic millimeter (65% neutrophils, 24% lymphocytes, and 11% monocytes), 1.3 g of protein per deciliter, 0.8 g of albumin per deciliter, 0.3 mg of bilirubin per deciliter (5.1 μmol per liter), 42 mg of triglycerides per deciliter (0.5 mmol per liter), 43 U of amylase per liter, and 396 U of lactate dehydrogenase per liter. Cytologic examination showed inflammatory cells, Gram's staining showed no organisms, and cultures grew few enterococcus colonies. Testing for lupus anticoagulant was negative, and levels of protein S, protein C, and antithrombin III were lower than normal.
CT scans of the chest after the intravenous administration of contrast material showed multifocal consolidation in the middle lobe of the right lung and bilateral collapse of the complete lower lobe and partial upper lobe, with bilateral layering of pleural effusions. CT scans of the abdomen and pelvis after the intravenous administration of contrast material showed large amounts of hemorrhagic ascites, persistent thickening of segments of the distal small-bowel wall, and unchanged lymphadenopathy. The liver and spleen showed evidence of hypoperfusion. Cultures of blood and urine were sterile. Testing of nasal and tracheal aspirates for influenza A (seasonal and pandemic) and B viruses, adenovirus, respiratory syncytial virus, and parainfluenza viruses 1, 2, and 3 was negative. Other test results are shown in Table 1.
On the second day, additional ascitic fluid was removed, and oliguric renal failure developed; a urine sediment revealed muddy brown casts. Continuous venovenous hemofiltration was begun.
During the next week, the patient required multiple pressors, mechanical ventilation, and frequent transfusions; broad-spectrum antimicrobial therapy was changed to vancomycin, metronidazole, ceftriaxone, peramivir, and micafungin and after 6 days to linezolid and meropenem. Stress doses of hydrocortisone were administered. The level of C-reactive protein was 28.2 mg per liter (reference range, <8.0 for inflammation); a test for antinuclear antibodies was positive (titer, 1:40 dilution), in a speckled pattern, and other tests for autoantibodies were negative; and there was no factor V Leiden or G20210A prothrombin gene mutation. Other test results are shown in Table 1. Pathological examination of the specimens from the other hospital showed transmural necrosis and acute inflammation of the bowel, a helminth in the bowel lumen, and a mesenteric lymph node with hemorrhage and acute inflammation, features consistent with acute hemorrhagic lymphadenitis. On the sixth day, hypotension worsened.
On the seventh day, laparotomy revealed ascites, blood clots in the pelvis, and a large infra-mesocolic retroperitoneal hematoma that extended to the bifurcation of the great vessels and involved the origins of the renal arteries. The bowel anastomoses were intact, with no evidence of ischemic viscera (Figure 1FIGURE 1
Intraoperative Photograph during Laparotomy.
). Two liters of ascites and the pelvic clots were removed, and the abdomen was left open for later closure. After the procedure, oxygenation and hypotension improved; the patient was weaned from pressors.
A diagnostic test result was received.
DIFFERENTIAL DIAGNOSIS
Dr. Mark S. Klempner: May we review the imaging studies?
Dr. Susanna I. Lee: Coronal reconstruction images from the abdominal CT performed at the other hospital, obtained after intravenous and oral administration of contrast material, show thickened loops of ileum, extending to the terminal ileum, with sparing of the colon. Numerous hyperenhancing lymph nodes of approximately 1 cm in the short-axis view are present in both the retroperitoneum and the small-bowel mesentery (Figure 2A and 2BFIGURE 2
Abdominal Images.
). There is also a large amount of ascites.
A chest CT scan obtained during the first 24 hours at this hospital, after the administration of intravenous contrast material, shows bilateral moderate-size layering of pleural effusions, with collapse of the lower lobes and partial collapse of the upper lobes. In the parts of the lungs that are aerated, there are patchy air-space opacities, which could represent either inflammation or infection. The abdominal CT scan obtained at the same time shows postoperative changes from the bowel resection but is otherwise unchanged.
Dr. Klempner: Four clinical features define the systemic inflammatory response syndrome (SIRS): fever or hypothermia, tachycardia, tachypnea, and leukocytosis or leukopenia.1 This patient fulfills three of the four criteria — tachycardia, tachypnea, and leukocytosis. The presence of SIRS conjures a differential diagnosis that includes noninfectious and infectious causes. We can use our understanding of the pathophysiology of this overwhelming and dysregulated inflammatory response as a framework for our differential diagnosis. Two different pathways may result in SIRS2,3 (Figure 3FIGURE 3
Pathways of Development of the Systemic Inflammatory Response Syndrome (SIRS).
). One is triggered by the release of endogenous cellular contents from damaged tissue or dying cells, known as damage-associated molecular patterns (DAMPs). The other is triggered by exogenous microbial exposure — so-called pathogen-associated molecular patterns (PAMPs). Both the endogenous DAMPs and the exogenous PAMPs are recognized by pattern-recognition receptors on cells of the innate immune system, including polymorphonuclear leukocytes and macrophages.
Damage-Associated Molecular Patterns
DAMPs trigger signaling cascades that result in the release of excessive inflammatory cytokines, which cause the physiological changes that we observe in this patient. DAMPs can be released from virtually any severely injured tissue or organ but are most commonly associated with injury to the pancreas, lungs, skin, soft tissues, or liver. Pancreatitis from alcohol, drugs, gallstones, or ischemia; inhalation or thromboembolic lung injury; burns, trauma, or surgery; and autoimmune diseases that are associated with vasculitis and tissue injury are the usual suspects that account for the release of DAMPs.4 In this critically ill woman, we must quickly consider endogenous DAMP-induced SIRS.
This patient has no history of hepatobiliary disease or alcohol or drug abuse. Since she initially had normal levels of liver and pancreatic enzymes and no liver or pancreas abnormalities on abdominal CT or at laparotomy, we can rule out pancreatitis or liver injury as the trigger for the SIRS. Similarly, there is no history of trauma, antecedent surgery, or inhalation of a toxic substance. Although she was taking an oral contraceptive and has a family history of hypercoagulable disorders, there is little in her history, physical examination, or laboratory-test or imaging results to suggest that thromboembolic disease accounts for SIRS caused by DAMPs of lung origin.
Systemic vasculitides can result in tissue damage sufficient to cause SIRS.5 This patient presented with acute symptoms and has little medical history, negative laboratory studies for autoantibodies, and findings on imaging studies and exploratory laparotomy that show inflammation and necrosis of mesenteric lymph nodes but no apparent vasculitis, all of which essentially rule out autoimmune vasculitis in this case.
Pathogen-Associated Molecular Patterns
Having ruled out endogenous causes of SIRS, we are left with the far more likely possibility that SIRS in this patient has been triggered by an exogenous microbial PAMP. If infection is documented, her presentation would be considered the sepsis syndrome.
We need to focus on clues that might suggest a specific agent as the microbial cause of septic shock in this patient. Her symptoms of rapidly worsening abdominal pain, nausea, and vomiting, as well as her distended and tender abdomen, point to an intraabdominal process. The imaging studies and findings at the time of the initial exploratory laparotomy also indicated an intraabdominal process. The unusual finding of mesenteric lymphadenitis with necrosis and hemorrhage allows us to focus our differential diagnosis on pathogens that are associated with intraabdominal infections and, more specifically, on those that cause mesenteric lymphadenitis.
Viral Infections
Viruses such as measles, Epstein–Barr virus (EBV), adenovirus, and the human immunodeficiency virus (HIV) can cause disseminated infection and result in mesenteric lymphadenitis.6 However, disseminated infection with measles, EBV, and adenoviruses usually occurs in early childhood, before specific immunity develops, and each of these viruses is associated with characteristic presenting symptoms and signs outside the gastrointestinal tract. These viruses often cause lymphadenopathy outside the abdomen and do not typically induce the sepsis syndrome. It is important to recognize these viruses as causes of mesenteric lymphadenitis, because they must be considered in the differential diagnosis of acute appendicitis in children. Primary HIV infection may be associated with diffuse lymphadenopathy, including mesenteric lymphadenopathy, but is not associated with a sepsis syndrome. The absence of a rash, pharyngitis, hepatitis, or an atypical lymphocytosis makes the diagnosis of acute HIV infection unlikely.
Bacterial Infections
Bacteria are much more likely than viruses to cause the combination of abdominal symptoms, mesenteric lymphadenitis and systemic inflammation or sepsis.
Campylobacter
Campylobacter, especially Campylobacter jejuni and C. coli, are foodborne or waterborne pathogens that typically cause diarrhea. Illness may start with acute abdominal pain and diarrhea; some patients have a 24-to-48-hour febrile prodrome consisting of fever and generalized aches before the onset of gastrointestinal symptoms. The spectrum of acute campylobacter-induced illness includes the “pseudoappendicitis syndrome,” which involves mesenteric lymphadenitis in the region of the terminal ileum and cecum. Patients with this syndrome may present with fever and abdominal pain but little diarrhea. Campylobacter can also cause colitis, with fever, abdominal pain, and bloody diarrhea, which can present the clinician with the possibility of inflammatory bowel disease, especially in a young person such as this patient. Our patient's gastrointestinal involvement resulted in necrotic bowel and not in diarrhea associated with ileitis or colitis, the mesenteric lymphadenitis was not in the right lower quadrant, and she had prominent septic physiology, which would be quite unusual with this syndrome.
Yersinia
Of the three pathogenic yersinia species, Yersinia enterocolitica and Y. pseudotuberculosis can cause a clinical syndrome dominated by fever, abdominal pain, and diarrhea. Both are transmitted by consumption of contaminated food or water. After ingestion, they invade the intestinal epithelium and localize to lymphoid tissue in the intestinal mucosa and in mesenteric lymph nodes. Because of their tropism for intraabdominal lymphoid tissue, both yersinia species are associated with the pseudoappendicitis syndrome and mesenteric lymphadenitis. In a classic outbreak of Y. enterocolitica infection associated with contaminated chocolate milk, almost half of the 95 affected persons underwent laparotomy for suspected acute appendicitis.7 Yersinia can also disseminate to involve almost any organ or tissue; persons with iron-overload syndromes are particularly susceptible to disseminated yersinia infection. However, except in these patients, sepsis is rare, mesenteric lymphadenitis is localized and not diffuse, and prominent ascites is rare. For these reasons, infection with yersinia is unlikely to explain this patient's presentation.
Mycobacteria
Both Mycobacterium tuberculosis infection and M. bovis infection can cause mesenteric lymphadenitis.8 Although M. tuberculosis is transmitted from person to person by inhalation and M. bovis by ingestion of contaminated dairy products, the resulting clinical syndromes can be indistinguishable, especially in extrapulmonary sites. The gastrointestinal tract or peritoneum can be the primary site of infection, although this site is more common with M. bovis. Although the so-called wet form of abdominal tuberculosis is associated with ascites, as was found in this patient, the clinical course of abdominal mycobacterial infection is rarely fulminant or associated with the sepsis syndrome. Necrosis of bowel is also rare, although ischemia may result in scarring and strictures. Mesenteric lymphadenitis is usually located in the right lower quadrant, and the pathological features usually include caseating granulomas and not the acute hemorrhage and inflammatory reaction observed in this patient. For all these reasons, abdominal mycobacterial infection seems unlikely in this patient.
Anthrax
Anthrax is caused by Bacillus anthracis and is associated with three clinical syndromes — cutaneous, inhalational, and gastrointestinal — depending on the site of entry of the spores. Gastrointestinal anthrax follows the ingestion of food — typically meat — contaminated with B. anthracis spores. 9 When ingested spores reach the stomach or bowel, they germinate on the gastrointestinal epithelium, causing ulceration, edema, and necrosis, and bacteria are transported to the lymphatic tissue in the bowel wall and mesenteric lymph nodes, where they cause hemorrhagic lymphadenitis. Clinical symptoms of fever, nausea, vomiting, and rapidly progressive severe abdominal pain usually begin within a few days after ingesting the spores. Large-volume ascites, due to the production of edema factor by the bacteria, is usually present. High-grade bacteremia and the features of sepsis may occur as part of disseminated anthrax bacilli originating in the gastrointestinal tract.
This case has all the clinical, laboratory, and initial pathological features of intestinal anthrax. It is worth noting that two of two bottles from the patient's initial blood culture were positive for gram-positive rods, which most likely were B. anthracis. Interpretation of this blood culture as a true positive is favored because of the clinical features of sepsis and the rapid outgrowth in both bottles. In this patient, the persistence of severe ascites and features of SIRS, despite effective antibiotic therapy, and multiple negative cultures of blood and ascites can most likely be attributed both to the effects of the anthrax toxin (a three-part protein, composed of protective factor, edema factor, and lethal factor), which does not directly interact with pattern-recognition receptors but causes cell death and the release of DAMPs,10 and to the presence of bacterial PAMPs in residual cell-wall components.11
Dr. Eric S. Rosenberg (Pathology): Dr. Cohen, would you tell us your clinical impressions when you saw this patient?
Dr. Louis Cohen (Medicine): The description of segmental necrotic bowel, hemorrhagic mesenteric lymph nodes, and the presence of gram-positive rods in the blood made us consider the diagnosis of anthrax. The patient's history of working on a farm and the report of a nonhealing leg wound with a possible eschar hinted at a potential portal of entry. Her presentation was strikingly similar to reports of gastrointestinal anthrax in the literature. However, it was difficult to reconcile a possible cutaneous infection many months before her presentation with gastrointestinal anthrax. Since she was a vegan and gastrointestinal anthrax is typically caused by ingestion of contaminated meat, it was unclear how she could possibly have acquired this infection.
The diagnostic procedure was further analysis of the isolate identified by blood culture.
CLINICAL DIAGNOSIS
Intestinal Bacillus anthracis infection with sepsis.
DR. MARK S. KLEMPNER'S DIAGNOSIS
Intestinal Bacillus anthracis infection with sepsis.
PATHOLOGICAL DISCUSSION
Dr. Mary Jane Ferraro: From the other hospital, we obtained the bacterial isolate that grew from the patient's blood cultures. The morphologic features of the colony were consistent with those of bacillus species. A Gram's stain revealed chains of large, gram-positive rods with subterminal spores, and the isolate was nonhemolytic on blood agar and nonmotile, features consistent with B. anthracis. The isolate was referred to the Massachusetts Department of Public Health State Laboratory Institute with a specific request to rule out this organism. Subsequently, B. anthracis DNA was detected by real-time polymerase chain reaction, and B. anthracis was also identified by the state laboratory with the use of standard biochemical methods. Susceptibility tests were performed at the Centers for Disease Control and Prevention (CDC). The isolate was found to be susceptible to cipro-floxacin, doxycycline, levofloxacin, penicillin, and tetracycline.
In our experience, bacillus species are not common contaminants of blood cultures, and it is important not to dismiss this genus as a common contaminant, such as might be the case for coagulase-negative staphylococci or diphtheroids.
Dr. Rosenberg: After the organism was identified, the slides and paraffin-embedded blocks of the intestinal and lymph-node specimens were sent to the CDC for further evaluation. Dr. Zaki, can you show us your findings?
Dr. Sherif Zaki: Histopathological examination of the small intestine showed transmural hemor-rhage, edema, and acute inflammation with necrosis (Figure 4A and 4BFIGURE 4
Histopathological Features of Specimens of Intestinal and Mesenteric-Lymph-Node Biopsies.
). Several nematodes seen within the lumen of the small intestine and appendix had prominent lateral alae and were morphologically consistent with Enterobius vermicularis. Examination of the lymph node revealed diffuse hemorrhage with acute inflammation and necrosis (Figure 4C). Numerous gram-positive rods in the wall of the small intestine (Figure 4D) and in the lymph node were identified on histochemical staining. The pathological differential diagnosis of an acute necrotizing and hemorrhagic enteritis and lymphadenitis with gram-positive bacilli would be necrotizing enteritis caused by Clostridium perfringens or gastrointestinal anthrax.
We performed immunohistochemical staining of the lymph node and small-bowel specimen with the use of two monoclonal antibodies directed against the capsule and cell wall of B. anthracis (Figure 4E and 4F). There were large amounts of bacterial antigens in both the small bowel and the lymph node. The large amount of antigen present was evidence of the remarkable bacterial load that is characteristic of this disease.
Dr. Rosenberg: Dr. Talbot, would you tell us how you think this patient may have become infected with anthrax?
Dr. Elizabeth A. Talbot: When the isolate identified from the blood culture was confirmed as B. anthracis, an immediate multijurisdictional investigation involving the New Hampshire and Massachusetts state health departments, the CDC, and the Federal Bureau of Investigation was begun to identify the source of anthrax and to ensure public safety. Hypotheses regarding the patient's source initially focused on her vegan diet and employment in organic farming and art. However, it was discovered that she recently had participated in a drum event in which drums made with animal hides were used. Human anthrax has previously been associated with animal-hide drums,12-14 and the timing of her attendance at the event and the onset of symptoms were consistent with the incubation period for gastrointestinal anthrax; therefore, the drums were thought to be the most likely source of her infection with anthrax.
Anthrax testing was performed at the event site and on the drums. Three samples grew B. anthracis: one from the event site and two from drums. The event site was quarantined, and additional testing yielded more positive samples. B. anthracis isolates from the patient, drums, and venue were found to be identical according to genotyping performed at the CDC. Using a variety of public health surveillance systems, we sought, but did not find, any additional human cases of anthrax. Decontamination of the site was initiated.
Drumming on an anthrax-contaminated animal-hide drum may have caused a brief aerosolization of anthrax spores. The patient may then have inhaled aerosolized spores into her oropharynx and then ingested them. It is also possible that spores had contaminated bread or water that she consumed at the event. This case is unique among previously reported cases of anthrax that were associated with animal-hide drums, because it is the first example of gastrointestinal disease and because of the minimal intensity of the patient's exposure.
Dr. Rosenberg: Dr. Stamm, will you tell us what happened to this patient?
Dr. Luisa M. Stamm (Infectious Disease): When the diagnosis of gastrointestinal anthrax was confirmed, the patient was already taking vancomycin and meropenem, which have some activity against B. anthracis. We added ciprofloxacin and clindamycin, which theoretically inhibits toxin formation. In collaboration with the CDC, she received anthrax intravenous immunoglobulin (pooled serum samples from vaccinated persons, which contained antibodies against the protective antigen of anthrax, which is required for the functioning of edema factor and lethal factor toxins). She was the fifth patient in the world to receive this treatment. By hospital day 24, after intravenous antibiotics had been administered for 14 days, the antimicrobial regimen was narrowed to ciprofloxacin. She continued to receive ciprofloxacin for nearly 60 days; near the end of the course she was switched to doxycycline because of side effects.
The patient was hospitalized for nearly 2 months. By day 9, sepsis resolved and she was removed from vasopressor support; continuous venovenous hemofiltration was discontinued the next day, and she was extubated on day 18. Her abdomen was closed on postoperative day 1 but she continued to have high-output ascites throughout the hospital course, with more than 50 liters of ascitic fluid drained from her abdomen. She had intermittent fever, for which she received empirical antibiotics, but B. anthracis was never isolated again.
The patient was discharged to a rehabilitation facility on hospital day 61. After 20 days of rehabilitation, she was discharged home. She continues to have ascites, nausea, vomiting, and abdominal pain, which are slowly resolving. The liver-function results remain elevated in a cholestatic pattern but are also slowly normalizing. She was malnourished and has lost 30 lb (13.6 kg). She is weak and deconditioned but is improving every day. In fact, she felt well enough to attend the conference today with her parents.
Dr. Klempner: Do we know anything about the origin of the drums?
Dr. Talbot: We have an itinerary for the 64 drums that were present at the time of the drumming event. The two drums that tested positive for B. anthracis were in this country for more than 2 years. One was a gift, and we do not know much about its history before it was given.
Dr. Arnold Weinberg (Infectious Disease): Dr. Klempner, how do we reconcile the fact that the more common effects of drum-related anthrax would be cutaneous or inhalational, rather than gastrointestinal?
Dr. Klempner: I suspect that the spores were either relatively large or clumped and were aerosolized and then swallowed.
ANATOMICAL DIAGNOSIS
Intestinal anthrax.
This case was presented at the Medical Grand Rounds, March 25, 2010.
Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.
No potential conflict of interest relevant to this article was reported.
We thank Marcia B. Goldberg, M.D., Cammie F. Lesser, M.D., Ph.D., and David R. King, M.D., for assistance in preparing the case history.