Abdominal Pain

The initial evaluation of a patient with acute abdominal pain should focus on identifying life-threatening causes and surgical emergencies such as  (Vasculitis) NEJM 2013
  • cholecystitis, 
  • aortic dissection, 
  • mesenteric ischemia, and 
  • bowel perforation
  • Other causes of acute abdomen include
    • appendicitis,
    • relatively benign illness (e.g., the passage of a kidney stone) or 
    • more ominous (e.g., luminal occlusion, thrombosis, or visceral or vessel rupture)
  • Other pearls
    • Gastroenteritis is unlikely, given the severity of the pain, but it should be considered routinely in a patient presenting with abdominal pain, nausea, and vomiting.

Murphy sign:  

The Rovsing sign (palpa- tion of the left lower quadrant resulting in pain in the right lower quadrant) 

Post-prandial Pain Mesenteric Ischemia NEJM 2016
  • Symptomatic Cholelithiasis
  • PUD 
  • Mesenteric Ischemia
  • Pancreatitis
  • Diverticular Disease
  • Gastric Reflux
  • Irritable Bowel Syndrome 
  • Gastroparesis
Building Differential DDx of And Pain: Key Principles
  • Acute Abdominal Pain Case 12-2014: (Acute Abdominal Pain) (Lead Poisoning)
    • Intrabdominal Causes of acute abdomen
      • More catastrophic causes 
        • gastrointestinal perforation including PUD with perforation or complications, 
        • intestinal infarction, 
        • ruptured abdominal aortic aneurysm
        • Mesenteric Ischemia with or without infarction
          • Look for the distribution of the abnormal bowel territory. Does it conform the vascular territory of the SMA or IMA etc.
          • Is  visualized portion of the superior and inferior mesenteric arteries patent 
        • Ischemic Colitis
        • Intestinal obstruction
          • Transition point is usually identified to suggest mechanical bowel obstruction. There is no free air or ascites. There is a chronic nonunited left pelvic fracture through which a loop of large bowel had herniated, without evidence of strangulation. The findings on CT performed during the patient's second admission are unchanged.
      • Other common Causes
        • appendicitis, 
        • diverticulitis, 
        • cholecystitis, 
        • pancreatitis, 
        • renal colic
        • inflammatory bowel disease
        • gastroenteritis
        • hepatitis
        • budd-chairi
    • Extrabdominal Cause of acute abdomen
      • colonic pseudo-obstruction, 
      • Acute porphyria
      • Lead poisoning 
  • Build DDx using Anatomy and Character of Pain
    • RUQ
      • Liver
      • GB and CBD
      • Kidney
      • Colon and Intestines
      • Duedenum
      • Skin and Nerves
      • Subcutaneous structures
    • RLQ
    • LUQ
    • LLQ
    • Suprapubic
    • Epigastric
    • Diffuse
    • But remember this: 
      • In patients presenting to an outpatient clinic with a new onset of abdominal pain, the location of the pain may or may not be predictive of the underlying pathologic features. In one study, sensitivity was high for pain in the epigastric region, indicating gastroduodenal processes; for right subcostal pain, indicating hepatobiliary diseases; and for mid-to-lower abdominal pain, indicating gynecologic diseases among women. This study suggests that we probably should not focus our differential diagnosis on organs in the left lower quadrant to the exclusion of other abdominal organs.
  • Acute vs Chronic Abdominal Pain
  • Acute Abdominal Pain (Medical or Surgical Emergencies)
    • 1) Acute Cholecystitis 
    • 2) Aortic Dissection
    • 3) Mesenteric Ischemia
    • 4) Bowel Perforation 

Case Examples of Abdominal Pain
Do not rely on the not so good test ...also, read Case 29-2010 below to not to rely on the not so good test to reach a diagnosis..

58 yo M is seen in the outside hospital for 1 day of diffuse abdominal pain that first started in the epigastric area. It has since then localized to the RLQ, is sharp, 9/10, no other radiation, movement makes it worse. Patient has had mild subjective fever and chills, Nausea, and 1 episode of non-bloody, non-bilious vomiting. He has never had any colonoscopy. CT non-contrast was done at the outside hospital and the official reading is "terminal ileum is inflamed, appendix is normal". Actual film is not available for review. Vitals and Labs are stable. WBC of 15 is the only finding. Patient is admitted to the floor. IV Cipro and IV Flagyl is started. Patient is NPO, and IV fluid is given. Next day, patient feels significant improvement, and is ready to eat. 
What is the next best step in the management. 
  • 1. Consult GI (for it could be first episode of CD, and needs steroids instead of abx)
  • 2. Continue abx for 10 days (abx is working, and is most likely infectious ileitis), and GI eval in 3-4 weeks for colonoscopy.
  • 3. Consult Surgery for it is an appendicitis, and CT from outside in not accurate. 
  • 4. CT with PO and IV contrast to better define the inflamed region. 
In this case, #4 was done. Showed 1.4 cm inflamed appendix. Patient is taken to the surgery for lab appendectomy. Gangrenous appendix is found. 
Teaching point: 
  • Stick to your history and exam. It is one of the valuable tool.
  • Do not rely on imaging read from outside hospital if you do not know the radiologist well. 
  • Do not rely on sub-optimal CT. CT without contrast cannot fully evaluate the abdominal pain of inflammatory origin. 
What would you have done if patient had CKD-3b. 
CT with PO contrast, and IV contrast with low dose IV contrast. Give fluid before and after CT. Give mucormyst to prevent contrast induced Nephropathy.

    Teaching Point: 
        i) In patient with unexplained abdominal pain, consider Vasculitis (PAN) as one of the DDx. 
        ii) PAN if untreated has poor prognosis (can also cause MI, stroke)
        iii) Inflammatory markers (ESR and CRP) may be high if suspected, C3, C4 if high. Consider UDS for cocaine
        iv) Can only involve one vessel or one organ, and have few aneurysm, making such a diagnosis difficult 
to reach 


Dr. Xavier: This 59-year-old man presented with months of recurrent nausea, vomiting, and abdominal pain. Duodenal-biopsy specimens reportedly showed active inflammation with villous blunting, crypt hyperplasia, and increased intraepithelial lymphocytes. Colonic-biopsy specimens showed an inflammatory infiltrate. Results of laboratory tests showed evidence of malabsorption, with folate and iron deficiency, mild anemia, and hypoalbuminemia; the patient had profound weight loss and malnutrition. Imaging scans and biopsy specimens showed no abdominal tumor.

Celiac Disease

Patients with celiac disease are usually younger than this one; however, an initial diagnosis after 60 years of age is not uncommon. Celiac disease is strongly associated with certain HLA types (HLA-DQ2 and HLA-DQ8) and is most common in persons of northern European descent. 

Classic gastrointestinal symptoms of celiac disease include those of malabsorption, such as steatorrhea, flatulence, and abdominal discomfort. Symptoms of celiac disease may range from fatigue and no gastrointestinal symptoms to profuse diarrhea with metabolic disturbances. Dermatologic manifestations include eczema and dermatitis herpetiformis, neither of which fits the description of this patient's skin lesions.

The sensitivity and specificity of IgA antibodies to tissue transglutaminase are greater than 94% in the absence of IgA deficiency; IgA deficiency can occur in up to 2% of persons with celiac disease. These negative results do not rule out a diagnosis of celiac disease, but they do reduce its likelihood. HLA testing could be considered if clinical suspicion for celiac disease is high despite negative serologic testing.

A biopsy specimen of the small bowel is a cornerstone of the diagnosis of celiac disease and typically reveals villous atrophy, crypt hyperplasia, increased intraepithelial lymphocytes, or a combination of these. There may also be an increase in intraepithelial lymphocytes in the colon. Although some features of this patient's small bowel–biopsy specimen resemble those of celiac disease, the findings are not specific for the diagnosis and may occur in cases of enteritis from other causes.

Adherence to a gluten-free diet should alleviate symptoms and signs of celiac disease.

Common Variable Immunodeficiency

A spruelike illness may occur in patients with common variable immunodeficiency (CVID). Patients with CVID have reductions in serum levels of IgG, IgA, IgM, or a combination of these. They also have poor responses to immunizations and often have recurrent infections, including sinopulmonary bacterial infections, opportunistic fungal infections, or protozoal infectionsThey may have seemingly paradoxical autoimmune manifestations, such as autoimmune cytopenias. Of patients with CVID, 20% have gastrointestinal manifestations (e.g., chronic giardiasis, spruelike illnesses, inflammatory bowel disease, protein-losing enteropathy, nonspecific malabsorption-like syndromes, or gastrointestinal lymphomas). Biopsy specimens of the small and large bowel may show pathological features that are indistinguishable from those of celiac disease. The diagnosis of CVID is usually made before the patient is 30 years of age.

Tropical Sprue

This patient was raised in a Caribbean country and had recently visited there. Tropical sprue, initially described by William Hillary in 1759, is endemic in many tropical regions, including the CaribbeanCharacterized by chronic diarrhea, malabsorption, and nutritional deficiencies of folate and vitamin B12, tropical sprue should be suspected in anyone who has lived for more than a month in a region where the disease is endemic. Symptoms may develop up to several years after emigration. On examination of biopsy specimens, tropical sprue mimics celiac disease. The cause is presumed to be infectious, and treatment with broad-spectrum antibiotics is usually curative. Patients with tropical sprue usually present with voluminous diarrhea, which our patient does not have. Although I cannot rule out tropical sprue in this case, I believe there is a more specific diagnosis that will fit this case better.

Crohn's Disease

Crohn's disease must be considered in this patient. This disorder affects the small bowel in 80% of afflicted patents, and the terminal ileum is the most commonly involved site. However, symptoms can occur anywhere from the mouth to the anus. Histopathological examination of mucosal-biopsy specimens may reveal a spectrum of severity, from increased intraepithelial lymphocytes to frank ulceration and inflammation, with architectural distortion and noncaseating granulomas. 

Although the patient's symptoms are compatible with a diagnosis of Crohn's disease, epidemiologic factors make this diagnosis unlikely. In the United States, the prevalence of Crohn's disease in the Hispanic population is one tenth that in the white population. The incidence of inflammatory bowel disease is increasing in the developing world; however, in the United States, persons who have lived in latitudes closer to the equator before the age of 30 years have a lower risk of the development of inflammatory bowel disease than those who have lived in more northern latitudes. Finally, this patient's dermatologic manifestations do not resemble the extraintestinal manifestations of inflammatory bowel disease (i.e., pyoderma gangrenosum and erythema nodosum).


An important clue that emerges from the patient's clinical history is the incidental finding of S. stercoralis in the jejunum 4 years earlier. S. stercoralis is endemic in the tropics and subtropics; in the United States, it is often diagnosed in recent immigrants or U.S. military personnel who have recently returned to the United States. The life cycle starts in the soil, where rhabditiform larvae develop into infectious filariform larvae that penetrate the skin, enter the systemic circulation, penetrate the alveolar spaces, are coughed up and swallowed, and enter the gastrointestinal tract. In the small intestine, the organism matures and releases eggs that develop into rhabditiform larvae, which are typically excreted in the stool. Autoinfection may occur, usually in immunocompromised persons, in which rhabditiform larvae mature into filariform larvae in the gut and penetrate through the wall of the large intestine or the perianal skin into the systemic circulation.

Most cases of strongyloidiasis are asymptomatic or cause only mild symptoms. An acute manifestation is duodenitis, which causes abdominal pain, nausea, vomiting, diarrhea, or a combination of these. Ground itch is a severely pruritic cutaneous manifestation of the diseaseChronic autoinfection may result in enterocolitis and malabsorption, with diffuse involvement of the upper gastrointestinal tract and the proximal large bowel. Dermal migration of the larvae may result in urticaria, a feature consistent with this patient's skin lesions, and areas of serpiginous erythema, known as larva currens. Pulmonary manifestations include dry cough and asthmalike symptoms. Rarely, a syndrome similar to Löffler's syndrome can be seen.

In this case, the recurring abdominal symptoms, the rash, and the results of examination of gastrointestinal-biopsy specimens obtained during endoscopic evaluation are consistent with a diagnosis of enterocolitis caused by strongyloides. I think it is unlikely that the patient cleared the initial infection. In addition, his low-grade fevers, weight loss, and profound malnutrition raise concern for a syndrome known as hyperinfection.

Strongyloides Hyperinfection Syndrome

Hyperinfection with S. stercoralis is the accumulation of a large burden of parasites during the autoinfection cycle. Parasites accumulate primarily in the colon, more in the right colon than in the left colon. The parasitic burden in the colon may be so high as to trigger mucosal compromise and sepsis caused by gram-negative rods. Major risk factors for hyperinfection are infection with human T-cell lymphotropic virus type I (HTLV-I) or the human immunodeficiency virus (HIV), iatrogenic immunosuppression, malignant tumors, and hypogammaglobulinemia. Eosinophilia may be absent, as it is in this case. Mortality associated with strongyloides hyperinfection is estimated to exceed 10%.

Of all the risk factors, infection with HTLV-I is the most likely in this patient, in view of his history. HTLV-I is endemic in the Caribbean, South America, southern Japan, south and central Africa, and the Middle East. Transmission typically occurs vertically from mother to child through breast-feeding but can also occur from sexual contact, blood transfusions, or intravenous drug abuse. Infection with HTLV-I promotes a type 1 helper T-cell (Th1) response (characterized by interferon-γ production and the promotion of a cellular immune response to intracellular pathogens), rather than a type 2 helper T-cell (Th2) response (characterized by the production of interleukins 4, 5, and 13 and IgE, facilitating a humoral immune response to extracellular pathogens); therefore, the host defenses against extracellular parasitic infections such as strongyloides are effectively down-regulated. For this reason, HTLV-I infection is also associated with treatment failure. Stool examination for ova and parasites can be insensitive in patients without hyperinfection, but organisms are usually detectable in patients with hyperinfection. The presence of filariform larvae and rhabditiform larvae in the stool is a clue that autoinfection has occurred, and a high parasite burden suggests hyperinfection. Serologic tests for anti-strongyloides antibodies can be helpful, but endoscopic biopsies can greatly assist in making the diagnosis. In patients with disseminated disease and pulmonary symptoms, the organism may be found in the sputum.

In summary, I believe the likely diagnosis in this case is S. stercoralis hyperinfection, in association with HTLV-I infection. If stool examinations for ova and parasites are negative, I would recommend performing endoscopic examination of the upper and lower gastrointestinal tract and obtaining biopsy specimens to look for the organisms.

Dr. Eric S. Rosenberg (Pathology): Dr. Gala, you saw this patient on the consultation service. What was your thinking when you saw him?

Dr. Manish K. Gala: I saw this patient on the second hospital day as part of the gastroenterology consultation team. Our differential diagnosis was similar to Dr. Xavier's. For the reasons he summarized, we doubted the diagnosis of celiac disease and were concerned about recurrent strongyloidiasis with autoinfection and possible hyperinfection. We recommended HLA typing for the HLA-DQ2 and HLA-DQ8 phenotypes, the absence of which would rule out celiac disease; protein electrophoresis and testing for blood immunoglobulin levels to rule out CVID; and examination of stool specimens for enteric pathogens, ova and parasites, and fecal fat. We scheduled upper and lower gastrointestinal endoscopy for the next possible day (fifth hospital day).

When tests with good sensitivity for your diagnostic consideration comes negative (IM tests)....and do not rely on the not so good test (CT with PO contrast without IV contrast) 


Two weeks before admission, fever developed in this patient, with no other reported symptoms. It would be essential for the admitting physician to explore symptoms that might localize a source for the fever.  After this smoldering illness, she had a relatively sudden onset of acute abdominal pain, flank pain, and foul-smelling urine. The initial examination revealed fever, mild tachycardia, tenderness of the sternum and left costovertebral angle, and left-sided abdominal tenderness, particularly in the left lower quadrant.

Abdominal Pain

In patients presenting to an outpatient clinic with a new onset of abdominal pain, the location of the pain may or may not be predictive of the underlying pathologic features. In one study, sensitivity was high for pain in the epigastric region, indicating gastroduodenal processes; for right subcostal pain, indicating hepatobiliary diseases; and for mid-to-lower abdominal pain, indicating gynecologic diseases among women. This study suggests that we probably should not focus our differential diagnosis on organs in the left lower quadrant to the exclusion of other abdominal organs.

However, we need to start somewhere, so we will begin with the most likely causes of pain in the left lower quadrant in a young woman: salpingitis, ectopic pregnancy, irritable bowel syndrome, inflammatory bowel disease, inguinal hernia, nephrolithiasis, and diverticulitis. The medical history indicates that the patient is not sexually active, making salpingitis and ectopic pregnancy unlikely. The absence of clinically significant bowel symptoms lessens the likelihood of irritable bowel syndrome or inflammatory bowel disease. An inguinal hernia would have to be incarcerated to generate the described degree of pain, and this should have been evident on examination. We are left with diverticulitis and nephrolithiasis, and diverticulitis is less likely than nephrolithiasis in a relatively young woman.

The presence of fever, tenderness in the left costovertebral angle, and left-sided abdominal tenderness, particularly in the left lower quadrant, suggests a recurrent ureteral stone with associated pyelonephritis, particularly in view of the patient's history. It would be helpful to know whether the pain was colicky, as it had been during her previous episodes of nephrolithiasis, and whether it had been present in a lesser degree during her 2-week illness. A kidney stone with pyelonephritis is unlikely because of the normal white-cell count and the absence of dysuria. The results of the initial urinalysis argue against a urinary tract infection, but complete obstruction of the left ureter needs to be ruled out, since obstruction might prevent the appearance of cells or bacteria in the urine. We will need to examine the initial abdominal CT scan carefully for evidence of a ureteral stone.

In patients presenting to the emergency department with acute abdominal pain, early abdominal CT scanning provides better diagnostic accuracy than does the standard practice of supine abdominal and erect chest imaging and less frequently misses unexpected and serious conditions. When different diagnostic strategies are compared in the evaluation of acute abdominal pain, CT has a sensitivity of 89% and a specificity of 77% in serious conditions. In cases in which ultrasonography is negative, ultrasonography followed by CT has the best sensitivity (94%) with the fewest missed serious conditions but has a slightly lower specificity (68%). Therefore, on the patient's presentation to the emergency department, it would be appropriate to obtain an abdominal CT scan.

Dr. Ashraf Thabet:  There is splenomegaly, with the spleen measuring 14.8 cm in the craniocaudal dimension. A mildly enlarged portacaval lymph node is present. There is no evidence of bowel obstruction, bowel inflammation, or large fluid collections.

Dr. Hunt: In this case, CT of the abdomen was very informative. Since the scan did not show diverticular disease, we could essentially rule out the diagnosis of diverticulitis. The scarring of the left kidney is consistent with previous upper urinary tract infection but does not explain the patient's acute symptoms on presentation.

The severity of this young woman's pain is of great concern, since it apparently was difficult to control with narcotic analgesic agents. We would be well advised to consider an admonition about severe abdominal pain from Cope's classic text, which says that the “majority of severe abdominal pains which ensue in patients who have been previously fairly well, and which last as long as six hours, are caused by conditions of surgical import.” Intra-abdominal processes that would generate very severe pain include bowel perforation, obstruction, or infarction; aortic dissection or rupture; and an acute inflammatory process (e.g., pancreatitis). There is no evidence for the first three processes on physical examination of the patient or on CT scans, and the normal white-cell and differential counts seem incongruous with the severity of pain and the implied urgency of an intra-abdominal process. The normal amylase and lipase levels and the normal appearance of the pancreas on the CT scan argue strongly against pancreatitis. Given the lack of intravenous contrast material on the initial CT scan, I am concerned that we may not be able to rule out a vascular process or an infarction of an intra-abdominal organ.

Splenomegaly and Lymphadenopathy

The finding of splenomegaly and lymphadenopathy in a patient with a protracted, intermittent fever expands our list of diagnostic considerations. Of these, the most likely in our patient would be infectious mononucleosis due to Epstein–Barr virus (EBV) or CMV. With this consideration, we should review the hematologic studies. The presence of atypical lymphocytes is intriguing, and the blood smear should be reviewed. If the percentage of atypical lymphocytes was substantially higher than reported, infectious mononucleosis would lead the diagnostic possibilities. Although the patient's elevated aminotransferase levels are evidence of hepatic involvement, the levels are lower than would be expected in a case of hepatitis A, B, or C, and viral hepatitis is unlikely to cause splenomegaly and lymphadenopathy. Bacterial infections are less likely in view of the normal white-cell count and the course of the illness. Infective endocarditis seems unlikely because of the absence of cardiac findings, the negative echocardiogram, and negative blood cultures. Since the patient had not traveled or been exposed to ticks or sick persons, mycobacterial, parasitic, rickettsial, and fungal illnesses are unlikely. Although systemic lupus erythematosus may be manifested in many ways, and we should keep this diagnosis on the list, there is little to suggest it. Lymphoma remains a possibility, and other hematologic processes may be associated with splenomegaly, but these would be less likely to account for the fever in the absence of a second process. Tests for EBV, CMV, and HIV should be performed in cases such as these. An antinuclear-antibody test would be reasonable, but we would need to interpret a positive result with caution.


We should briefly consider the anemia that worsened during the early hospital stay, coincident with a rising bilirubin level and an elevated lactase dehydrogenase level. The direct bilirubin level was more than 50% of the total, which argues for hepatic or post hepatic processes as the major cause of the patient's hyperbilirubinemia. The elevated red-cell distribution width combined with a low mean corpuscular volume prompts consideration of a relatively short list of causes, including iron deficiency, β-thalassemia, anemia of inflammation (in some patients), deficiency in glucose-6-phosphate dehydrogenase, and fragmented red cells. Although it is likely that several processes contributed to her anemia, the available iron studies suggest iron deficiency, the history suggests the possibility of anemia of inflammatory disease, and the lactase dehydrogenase level suggests possible coincident hemolysis.

By day 3 of the patient's hospitalization, the abdominal pain was localized to the left upper quadrant and left flank. This refocuses our differential diagnosis for the cause of her abdominal pain. Processes that have not yet been eliminated from our differential diagnosis include gastric ulcer, gastritis, splenic abscess, and splenic infarction. Intrathoracic or systemic processes might cause pain in the left upper quadrant and must be kept in mind, but knowing that the patient has splenomegaly should focus our attention on that organ. Repeat CT imaging after the intravenous administration of contrast material would be appropriate.

On day 5, the results of a number of diagnostic studies became available and should refine our working differential diagnosis. The negative tests for CMV, hepatitis, and antinuclear antibodies effectively reduce if not eliminate the associated diagnoses from consideration. My leading consideration is infectious mononucleosis, but the heterophile-antibody test was negative. The sensitivity of heterophile-antibody testing in the diagnosis of infectious mononucleosis ranges from 81 to 95%, with excellent specificity. Early in the course of illness, the heterophile-antibody test may be negative. At this point in the course of the patient's illness, I think that infectious mononucleosis remains the leading diagnostic consideration, despite the negative heterophile-antibody test.

Dr. Thabet: A subsequent CT scan of the abdomen and pelvis, obtained after the administration of intravenous and oral contrast material, shows persistent splenomegaly  In addition, multiple wedge-shaped areas of hypoenhancement are shown in the periphery of the spleen, features that are consistent with splenic infarcts. Other findings were unchanged.

Dr. Hunt: There are many causes of splenic infarction. and it is tempting to broaden the differential diagnosis again to include other causes, but I think this would be an error. Before extensive and expensive additional testing is ordered, we should determine whether splenic infarction is consistent with our leading diagnosis.

Splenic Infarction

The majority of patients with splenic infarction present with pain in the left upper quadrant or left flank or both, although a surprising percentage of patients are asymptomatic. The majority of patients with splenic infarction have elevated lactate dehydrogenase levels, and about a quarter of them have fever and chills. Some patients have splenic infarction with an infectious cause, such as EBV. An understanding that infectious mononucleosis has been associated with splenic infarction should prompt repeated or more sensitive testing for mononucleosis before alternative diagnoses are pursued.

Infectious Mononucleosis

If the diagnosis is infectious mononucleosis, does this account for the other features of this patient's illness? Elevation of aminotransferase levels occurs in 50 to 80% of patients with mononucleosis and in up to 3% of patients with hemolytic anemia. In one series, anticardiolipin antibodies, noted in this patient, were detected in the serum of 30% of patients with infectious mononucleosis. Furthermore, splenic infarction has been associated with anticardiolipin antibodies and infectious mononucleosis. 

Diagnosing this patient's illness is challenging and underscores the importance of appropriately framing the most prominent features of the illness, carefully reviewing all the imaging and laboratory data, and focusing further testing on the diagnoses of greatest clinical probability. I think that a diagnosis of infectious mononucleosis could have been made in this case by no later than the third hospital day, on the basis of a subacute illness in a 30-year-old woman with fever, splenomegaly, lymphadenopathy, hepatitis, and a normal white-cell count with evolving lymphocytosis, despite the negative heterophile-antibody test. Testing for antibodies to EBV-specific viral capsid antigen and EBV nuclear antigen proteins should be performed; the presence of IgM antibodies to EBV-specific viral capsid antigen would be diagnostic.

When 1, or even 2 biopsy and pathology findings are not enough...


Dr. Vijay Yajnik: This 80-year-old man with a complex medical history presented with a several-year history of intermittent abdominal pain, diarrhea, gastrointestinal bleeding, anorexia, and weight loss. Despite an extensive workup and multiple hospital admissions, a unifying diagnosis remained unknown. I will focus my differential diagnosis on several features of this case, including the findings on FDG-PET–CT (i.e., FDG uptake in several lung nodules, the stomach, and the ileocecal region), the identification of an ileocecal mass, an 8-mm induration at the site of a tuberculin skin test, and a positive interferon-γ release assay for Mycobacterium tuberculosis. Given this patient’s age and the constellation of findings, we need to consider processes that are likely to cause disease in the ileocecal region of the bowel, including cancer, Crohn’s disease, and infection.


Colon or Small-Bowel Cancer

Cancers of the colon and small bowel are relatively common in elderly patients and appear as a positive finding on PET. On endoscopy, the lesion often appears to be fungating. However, with an adequate biopsy specimen, such tumors are easy to diagnose on histologic examination. The tumors progress to invade local structures and then metastasize to the liver. Granulomas, which were described in this patient’s initial pathology report, are not typically seen. Furthermore, this patient’s lesion remained relatively stable on multiple imaging studies that were obtained over time, and there was no evidence of metastatic spread to the liver. For all these reasons, the diagnosis of colon or small-bowel cancer is unlikely in this case.

Carcinoid Tumor

Could this patient have a carcinoid tumor? This type of tumor can involve the ileocecal valve and appendix and may cause the carcinoid syndrome if the disease metastasizes to the liver. The endoscopic appearance of a carcinoid tumor can be either a submucosal prominence or an ulcerative lesion. On rare occasions, this tumor may cause a fungating mass lesion. Midgut carcinoid tumors are desmoplastic and, as a result, cause inflammation that leads to fibrosis and kinking of the mesenteryHowever, carcinoid tumors are easy to diagnose on histologic examination, and granulomas are not a typical pathological finding.

Primary Gastrointestinal Lymphoma

Primary gastrointestinal lymphoma may also occur in the region of the ileocecal valve and the appendix. Lymphomas are low-grade lesions with FDG uptake that can be stable for months; these features are consistent with those seen in this patient. On endoscopy, the lesions can be ulcerative in the upper tract and polypoid in the small bowel and colon and may even appear to be fungating. This patient had both an ulcerative lesion on upper endoscopy and a fungating lesion on colonoscopy. However, I suspect that the gastric ulcer was a result of chronic illness and that this process was distinct from the ileocecal mass. Although primary gastrointestinal lymphoma is a consideration in this patient, granulomas are typically seen in patients with Hodgkin’s disease and not in those with primary gastrointestinal lymphoma; therefore, this diagnosis is unlikely.

Crohn’s Disease

Throughout the case presentation, the diagnosis of inflammatory bowel disease was considered and the patient was offered therapy for Crohn’s disease on several occasions. Although Crohn’s disease can occur anywhere throughout the gastrointestinal tract, it commonly affects the ileocecal valve and terminal ileum. Crohn’s disease can be manifested at any age, including in an 80-year-old patient. However, stomach lesions are rare in elderly patients with Crohn’s disease, but focally enhancing gastritis is described in young patients with Crohn’s disease. Also, Crohn’s disease is associated with FDG uptake. On endoscopy, the lesion is predominantly ulcerative, but polypoid lesions are seen with expansion of lamina propria and lymphoid structures; a mass that appears to be fungating is rare. On histologic examination, granulomas are seen in 20 to 25% of cases and the disease can be stable for months; these features are consistent with the features of this patient’s presentation.


Bacterial, fungal, and mycobacterial infection can occur at the ileocecal valve, cecum, and terminal ileum. Bacterial and fungal infections are typically acute and cause progressive disease. However, similar to Crohn’s disease, intestinal tuberculosis can be indolent and cause chronic disease that remains stable for months; these features are consistent with those seen in this patient. Intestinal tuberculosis can be manifested at any age and is associated with FDG uptake; the endoscopic appearance is relatively indistinguishable from that of Crohn’s disease. Ulcerative, polypoid, and fungating lesions can be present. Granulomas are seen in 20 to 25% of cases. The mycobacterial load is low, so staining for acid-fast bacilli is often negative.

I suspect that a clue in this case was the positive interferon-γ release assay for M. tuberculosis that had been reported 11 months before this admission. On presentation, the patient had progressive anemia and leukocytosis. Both the erythrocyte sedimentation rate and the C-reactive protein level were high, and an abdominal CT scan that had been obtained 11 weeks before admission showed progression of inflammatory changes in the ileocecal area, with circumferential wall thickening and mesenteric fat stranding. Taken together, these findings are inconsistent with the diagnosis of primary epithelial, stromal, or neuroendocrine cancer or lymphoma. Other possible causes, such as gastrointestinal amyloid, Whipple’s disease, occlusive and nonocclusive mesenteric ischemia, and stromal tumors, typically are not associated with an inflammatory mass. The mesenteric fat stranding is suggestive of inflammation, and Crohn’s disease and intestinal tuberculosis are the most likely possibilities.

Crohn’s Disease versus Intestinal Tuberculosis

Does this patient have Crohn’s disease or intestinal tuberculosis?  Misdiagnosis of Crohn’s disease in a patient with intestinal tuberculosis would result in treatment with glucocorticoids and biologic agents, which then has the potential to cause disease progression that leads to increased morbidity and mortality. Misdiagnosis of intestinal tuberculosis in a patient with Crohn’s disease would lead to prolonged antitubercular therapy and delay the necessary immunosuppression required to induce disease remission.

Both diseases have an insidious onset. Diarrhea, hematochezia, and extraintestinal manifestations are more common in patients with Crohn’s disease. Intestinal tuberculosis can target extrapulmonary sites in a manner that resembles the classic extraintestinal manifestations of Crohn’s disease, such as reactive arthritis, erythema nodosum, and uveitis. Ascites and fever are more commonly seen in patients with intestinal tuberculosis. Both diseases involve the ileum and colonic segments of the bowel. Isolated involvement of the terminal ileum is commonly seen in patients with Crohn’s disease, whereas involvement of the ileocecal area and a patulous ileocecal valve is seen in patients with intestinal tuberculosis. In patients with Crohn’s disease, mucosal injury has a cobblestone appearance with aphthous and longitudinal rake ulcers, whereas in patients with intestinal tuberculosis, the ulcers are transverse in orientation. Furthermore, the granulomas associated with intestinal tuberculosis are more frequent and confluent and larger than those associated with Crohn’s disease. Tissue samples are positive for acid-fast bacilli in only 25 to 30% of cases of intestinal tuberculosis. The use of molecular techniques, such as polymerase-chain-reaction (PCR) assays of fresh biopsy specimens, can improve the diagnostic yield.

On admission, this patient was ill but did not have diarrhea, hematochezia, extraintestinal manifestations, or a fistula, features that are commonly seen in patients with Crohn’s disease. An inflammatory mass was present on endoscopy, and examination of the biopsy specimen revealed granulomas without visible acid-fast bacilli. However, these findings do not rule out the diagnosis of intestinal tuberculosis, because detection of acid-fast bacilli has poor sensitivity, presumably due to low organism load. The patient is originally from Southeast Asia, where tuberculosis is endemic, and an interferon-γ release assay for M. tuberculosis was positive, thus indicating previous exposure. Tuberculosis can be dormant in patients for several decades, and reactivation can occur in patients with immunocompromise, long-term use of glucocorticoids, diabetes, renal failure, and cancer. This patient had several of these risk factors, including glucocorticoid use, renal failure, and diabetes. Taken together, these findings favor a diagnosis of intestinal tuberculosis. Given the high likelihood for tuberculosis, I would perform a repeat colonoscopy with a biopsy and send the tissue for histologic examination, culture, and PCR assay for tuberculosis.

Dr. Virginia Pierce (Pathology): Dr. Khalili, what was your impression when you initially evaluated this patient?

Dr. Hamed Khalili: We met this patient on the second hospital day to help with management of his bloody diarrhea and abdominal pain. He had previously received a diagnosis of Crohn’s ileocolonic disease; however, a number of atypical features — including his age, country of origin, and previous endoscopic findings — were not entirely consistent with Crohn’s disease. Nevertheless, noncaseating granulomas that had been noted in previous biopsy specimens were strongly suggestive of Crohn’s disease. Since he was an elderly man originally from Southeast Asia and had a previously positive interferon-γ release assay for M. tuberculosis, we considered the diagnosis of intestinal tuberculosis. Because of the similarities in clinical and endoscopic findings between intestinal tuberculosis and Crohn’s disease, it was particularly important to evaluate the patient more extensively for intestinal tuberculosis. Given his coexisting conditions, we also considered the possibility of ischemic colitis. However, we thought that diagnosis was unlikely because of the location of involvement in the bowel. 

Acute Midabdominal Pain (Acute Mesenteric Vein Thrombosis) Mayo Clic Proc 2014

Peptic ulcer disease is an important cause of epigastric pain and can lead to severe diffuse abdominal pain if perforation occurs. However, one would expect peritoneal signs in the acute phase due to chemical peritoneal irritation, followed by evidence of infection as a secondary bacterial peritonitis develops. Appendicitis is an important cause of midabdominal pain that can often start in the periumbilical area before localizing to the right lower quadrant. However, in the case of a ruptured appendix, one would also expect peritoneal signs. Acute cholecystitis is a common cause of abdominal pain. However, it typically occurs in the right upper quadrant with radiation toward the right shoulder. It is often associated with recent fatty food intake and nausea. Because our patient had the classic finding of pain out of proportion to the abdominal examination findings that is seen in various intestinal ischemic syndromes, acute MVT is the most likely cause of his symptoms. Abdominal aortic aneurysm is also an important cause of midabdominal pain, but the pain associated with abdominal aortic aneurysm often radiates to the back. Further- more, this patient lacks vascular risk factors or a smoking history, making this diagnosis less likely.

Computed tomographic angiography of the abdomen and pelvis with intravenous contrast is the ideal test for diagnosis of MVT because it can quickly visualize the bowel, surrounding structures, and vasculature. It is highly sensi- tive and specific up to 93% and 96%, respec- tively, in diagnosing mesenteric ischemia and can provide direct visualization of the thrombus. 

Abdominal radiography with an upright view that includes the diaphragm would be an ideal test to quickly identify a perforated viscus by revealing free intraperitoneal air. However, in this patient who lacks peritoneal signs, this test would not be helpful in identi- fying the cause of his pain. Furthermore, although acute MVT leading to intestinal ischemia may exhibit a characteristic thumb- print of mucosal edema or pneumatosis intes- tinalis, these findings are nonspecific, and further evaluation would be needed. (In essence, do not do Xray abdomen)

Doppler USG have its limitations as well, hence CTA is the best test to go for..


Unless it is in your DDx, you can never make a diagnosis...


Dr. Lawrence S. Friedman: This 59-year-old man presented with an acute illness characterized by epigastric pain, which was followed by more diffuse abdominal pain that worsened with eating. He also had an unusual constellation of symptoms that raises the possibility of a single underlying disease, and disorders that do not account for many of the features of this case can be easily ruled out.

Intraabdominal Causes of Acute Abdominal Pain

Acute abdominal pain has a broad differential diagnosis that includes both intraabdominal and extraabdominal causesLife-threatening intraabdominal catastrophes, such as gastrointestinal perforation, intestinal infarction, and a ruptured abdominal aortic aneurysm, cause illness within minutes or hours and can be ruled out in this case by the length of the disease course and the lack of characteristic laboratory and imaging findings. Mesenteric ischemia without infarction can develop over a period of days, and abdominal pain that is disproportionate to the level of abdominal tenderness (as in this case) may be suggestive of mesenteric ischemia without infarction, particularly when the cause is mesenteric-vein or arterial thrombosis and when features of infarction (e.g., acidosis, an elevated lactate dehydrogenase level, and hyperamylasemia) are absent. In this case, a diagnosis of mesenteric ischemia is difficult to reconcile with the findings of a new anemia without clear evidence of gastrointestinal bleeding, the hyponatremia in the absence of intestinal infarction, and the absence of risk factors for mesenteric ischemia. A diagnosis of ischemic colitis in the absence of hematochezia would be unusual, and the findings are inconsistent with volvulus or an incarcerated hernia. Colonic obstruction, perhaps due to a neoplasm, has been ruled out by the CT findings; moreover, results of a colonoscopy had been normal 3 years earlier.

Common causes of acute abdominal pain — appendicitis, diverticulitis, cholecystitis, pancreatitis, and renal colic — as well as inflammatory bowel disease can be ruled out by the imaging and laboratory findings in this case. A bleeding peptic ulcer had been considered but was not confirmed on an upper endoscopic examination; also, the drop in hematocrit was not accompanied by gastrointestinal bleeding. Gastroenteritis was also suspected, but it was not consistent with this patient's clinical course.

The elevated blood levels of aminotransferases raise the possibility of hepatitis, but the severe pain and associated extraintestinal and laboratory features are not typical of the common hepatitis viruses or of a drug-induced hepatitis. Other infections that may involve the liver and have protean extrahepatic manifestations, such as visceral larva migrans, are unlikely because of the absence of eosinophilia, pneumonitis, and other characteristic features. This patient could have the complications of long-term alcohol use and the laboratory abnormalities associated with alcoholic hepatitis, but in that case, the blood level of aspartate aminotransferase would be higher than the blood level of alanine aminotransferase; also, in light of the severity of symptoms, evidence of hepatic dysfunction would be expected. The age of the patient makes Wilson's disease and other metabolic disorders, such as hereditary tyrosinemia, unlikely. Furthermore, the symptoms are inconsistent with nonalcoholic fatty liver disease and the Budd–Chiari syndrome. The elevated aminotransferase levels may be a nonspecific component of a systemic process.

Extraabdominal Causes of Acute Abdominal Pain

The constellation of features suggests that an extraabdominal cause of acute abdominal pain is most likely. Most entities in this category can be easily ruled out, because the diagnosis must explain the colonic pseudo-obstruction, the acute anemia without apparent gastrointestinal bleeding, and the acute hyponatremia, which appears to be caused by the syndrome of inappropriate secretion of antidiuretic hormone (SIADH).

Acute porphyria should be considered in a patient with acute, generally recurrent, severe abdominal pain that does not have a clear explanation. Features of this case that are consistent with a diagnosis of acute porphyria include severe neuropathic abdominal pain, nausea, stress or restlessness, pain in the arms and legs, constipation, colonic pseudo-obstruction, tachycardia, episodic hypertension, and SIADH; however, the characteristic dark urine was not present. Porphyrias are generally inherited and are caused by a deficiency in the activity of one of the enzymes required for normal heme synthesis. 

Microcytic Anemia NEJM 2014

They are categorized according to clinical presentation (neurovisceral or cutaneous) and according to the principal source of overproduction of porphyrins and porphyrin precursors (typically, the bone marrow or the liver)Patients with an acute neurovisceral porphyria, most commonly acute intermittent porphyria, may present with colicky abdominal pain, often in the lower abdomen. Patients with hereditary coproporphyria and those with variegate porphyria may present with either neurovisceral features or cutaneous features. Porphyria can be triggered by starvation, a negative energy balance, the use of drugs or alcohol, smoking, infections, and other forms of stress. This patient is 59 years of age, and a first attack of porphyria at this age would be unusual. Alcohol could have precipitated an attack, but pirbuterol is not included on the American Porphyria Foundation list of drugs that provoke attacks of acute porphyria or on other similar lists.

Causes of Dysgeusia and Basophilic Stippling

This patient could have had acute porphyria, but two findings that must be explained by the diagnosis are dysgeusia and basophilic stippling, and these are not known features of porphyria. Common causes of dysgeusia are chemotherapeutic agents, such as cyclophosphamide and cisplatin, and other drugs, such as albuterol, histamine H1-receptor antagonists, penicillamine, metronidazole, and boceprevir; it is also associated with conditions caused by exposure to pesticides and other toxins, such as lead poisoning, as well as zinc deficiency and xerostomia. Dysgeusia is a known side effect of pirbuterol.

Basophilic stippling is a hallmark of sideroblastic anemia and of lead poisoning, although it is not a constant feature of the latter. It is also seen in patients with arsenic poisoning, some thalassemias, a deficiency of erythrocyte pyrimidine 5'-nucleotidase, or thrombotic thrombocytopenic purpura.

Dysgeusia and basophilic stippling were distinct and prominent features in this case, although it is possible that nausea may have been perceived as altered taste. Arsenic poisoning causes a garlicky odor in the breath instead of true dysgeusia, and it is characteristically associated with severe diarrhea and cardiovascular symptoms, findings that were not seen in this case.

Lead Poisoning

I believe the most likely diagnosis in this case is lead poisoning, which explains all the clinical, laboratory, and imaging features, including the abdominal pain (“lead colic”), nausea, dysgeusia, constipation, colonic pseudo-obstruction, joint and muscle pain, behavioral and cognitive changes, acute anemia, basophilic stippling, SIADH, and decline in the blood level of phosphorus (which may be due to renal phosphate wasting). Lead lines, or bluish pigmentation at the gum–tooth line caused by a reaction of lead with dental plaque, are not a reliable indicator of acute lead poisoning and were absent in this patientDeposition of lead in bones may be seen with long-term exposure, as may hypertension and neuropsychiatric effects. The diagnosis of lead poisoning is confirmed by measuring the blood lead level; a level of 10 μg per deciliter or higher is considered elevated in adults. The level may be higher than 100 μg per deciliter in patients with acute lead poisoning, which is much less common than chronic lead poisoning. Heme synthesis is impaired when 5-aminolevulinic acid (ALA) dehydratase is 80 to 90% inhibited; this occurs at a blood lead level of approximately 55 μg per deciliter.  Because of the inhibition of ALA dehydratase and the overproduction of ALA, patients with lead poisoning or with hereditary tyrosinemia type 1 classically present with features that are similar to those of patients with acute porphyria. Indeed, plumboporphyria, a porphyria that is caused by a deficiency of ALA dehydrates, is named for “plumbum” (Latin for “lead”) because symptoms of the condition mimic those of lead poisoning, but plumboporphyria is rare and is generally reported in children. An elevated urine or blood level of ALA is often helpful in establishing a diagnosis of lead poisoning, although patients with lead poisoning and patients having an attack of acute porphyria would both be expected to have this laboratory abnormality. However, only patients with acute porphyria would also have an elevated porphobilinogen level. Free erythrocyte protoporphyrin and zinc protoporphyrin levels, which show the effect of lead on hemoglobin synthesis, can also be used as indicators of lead exposure over the preceding 3-month period.

Mesenteric Ischemia NEJM 2016


Mesenteric Circulation

The mesenteric circulation is extremely complex. Three primary vessels — the celiac artery, superior mesenteric artery, and inferior mesenteric artery interconnect through collateral networks between the visceral and nonvisceral circulations. These interconnections ensure that the loss of a single vessel does not lead to catastrophic malperfusion of the viscera.

The acute occlusion of a single vessel (typically the superior mesenteric artery) in acute mesenteric ischemia can result in profound ischemia caused by the loss of blood flow through this key vessel and its collateral vascular network. In contrast, in patients with chronic mesenteric ischemia, additional collateral networks develop over time; symptoms often do not appear until occlusion of two or more primary vessels occurs.

Causes of Altered Circulation and Mechanism of Injury

The causes of altered mesenteric circulation are themselves often the result of obstruction or diminished blood flow, with a resulting decrease in oxygen delivery to a level that is insufficient to meet the metabolic needs of the visceral organs. Vasodilatation is the initial response, but prolonged ischemia leads to vasoconstriction, which can persist even after intestinal blood flow returns to normal.This early injury primarily affects the intestinal mucosa and submucosa and potentially impairs mechanisms that prevent the translocation of bacteria from the intestinal lumen.

This sequence of events can result in the activation of systemic inflammatory pathways and ultimately in worsened vasospasm, further regional ischemia, and more extensive injury to the bowel wall. Without intervention, the damage can progress to full-thickness injury, infarction, and death.


History and Physical Examination

Early attention to the details of the patient’s history and to findings on examination that indicate the presence of mesenteric ischemia is critical for timely diagnosis and treatment. In contrast to other vascular disorders, mesenteric ischemia primarily affects women; more than 70% of persons with this disorder are female.

The physician should assess the patient’s records and the results of the examination for any evidence of other atherosclerotic and vascular diseases, including peripheral artery, cerebrovascular, coronary artery, and renovascular disease. In addition, other pulmonary and cardiovascular conditions must be identified and managed, since they are often coexisting conditions in patients with mesenteric disease and they may limit the available options for revascularization.

Manifestations of Acute Mesenteric Ischemia

Patients with acute mesenteric ischemia may initially present with classic “pain out of proportion to examination,” with an epigastric bruit; many, however, do not. Other patients may have tenderness with palpation on examination owing to peritoneal irritation caused by full-thickness bowel injury. This finding may lead the physician to consider diagnoses other than acute mesenteric ischemia. In a patient with abdominal pain of acute onset, it is critical to assess the possibility of atherosclerotic disease and potential sources of an embolus, including a history of atrial fibrillation and recent myocardial infarction. During the examination, the patient’s description of the history and symptoms can be unclear because of changes in mental status, particularly if he or she is elderly.

Differentiation between arterial and venous obstruction is not always simple; however, patients with mesenteric venous thrombosis, as compared with those with acute arterial occlusion, tend to present with a less abrupt onset of abdominal pain. Risk factors for venous thrombosis that should be evaluated include a history of deep venous thrombosis, cancer, chronic liver disease or portal-vein thrombosis, recent abdominal surgery, inflammatory disease, and thrombophilia.

Manifestations of Chronic Mesenteric Ischemia

Patients with chronic mesenteric ischemia can present with a variety of symptoms, including abdominal pain, postprandial pain, nausea or vomiting (or both), early satiety, diarrhea or constipation (or both), and weight loss. A detailed inquiry into the abdominal pain and its relationship to eating can be enlightening. Abdominal pain 30 to 60 minutes after eating is common and is often self-treated with food restriction, resulting in weight loss and, in extreme situations, fear of eating, or “food fear.” 

An extensive gastroenterologic workup, possibly including cholecystectomy and upper and lower endoscopy — tests that are often negative in patients with chronic mesenteric ischemia — is generally carried out before the diagnosis is made. An important distinction is that many of these alternative processes do not involve weight loss, whereas it is common in cases of mesenteric ischemiaSince older age and a history of smoking are common in these patients, cancer is often considered, and concern about it may delay the identification of chronic mesenteric ischemia. Nonetheless, particularly in the case of elderly women with a history of weight loss, dietary changes, and systemic vascular disease, chronic mesenteric ischemia must be seriously considered and evaluated appropriately.

Laboratory Studies

  • assessment of fluid, electrolyte, and acid–base status and evaluation for infection. 
  • Many patients present with acidosis due to dehydration and decreased oral intake. However, lactic acidosis often indicates at least segmental, severe ischemia or irreversible bowel injury. It is not helpful to wait for evidence of increasing serum lactate levels to proceed with further testing; ideally, in fact, intervention would occur in patients with acute mesenteric ischemia before lactic acidosis develops, with the goal of saving additional intestine from full-thickness injury. A left shift in the ratio of immature to mature neutrophils or an elevated white-cell count may indicate full-thickness injury to the bowel wall or ischemia with bacterial translocation.
  • Serum biomarkers have not proved to be as valuable for the early detection of acute mesenteric ischemia as was initially hoped. Despite the many investigations conducted to date, no clinically useful biomarkers have been identified, probably owing to the hepatic metabolism of complex proteins secreted by the intestine. Tests for markers of nutritional status, such as albumin, transthyretin, transferrin, and C-reactive protein, are the only studies of value in cases of chronic mesenteric ischemia, since they can be used to assess the degree of malnutrition before revascularization is undertaken.


Dr. Daniel P. Hunt: This relatively young man had had several illnesses during the few years before his current presentation, and he presented with abdominal pain that apparently resolved fairly quickly.

In drawing up a list of the patient's symptoms, signs, and history, I noticed that he had very few abnormal laboratory-test results. Therefore, the differential diagnosis must focus on the lymphadenopathy detected on examination, the abnormal findings seen on CT scans, and the history.

At this point, clinical instinct should lead us to carefully analyze and expand the history and to review the chest radiograph with an expert radiologist. I was able to do the latter in preparing for this case discussion, but of course I did not have access to the patient's record and I could not communicate with the patient to expand the history. Thus, I will list questions that might have been helpful to ask in an attempt to arrive at a unifying diagnosis, and then I will consider the history that was provided, making a few assumptions about its relevance to the current presentation.


We need to ask a few questions about this patient's history of psoriasis. What was the extent of involvement? Had there been joint involvement? How had the illness progressed over the years? Which treatments had he received? Had he received biologic agents? How did the psoriasis affect his life? We should ask these questions to estimate the potential for illnesses associated with psoriasis or long-term consequences. Associated illnesses include psoriatic arthritis, cardiovascular disease, depression, anxiety, nonmelanoma skin cancer, and lymphoma. I am particularly concerned about lymphoma in this patient with widespread lymphadenopathy and abnormal findings in the spleen, because a number of studies suggest that patients with psoriasis have an increased risk of lymphoma. Because of the potential association between psoriasis and lymphoma, we need to include lymphoma in our differential diagnosis.

Erythema Nodosum

Next, we should investigate the patient's history of erythema nodosum. At the time of the episode, was another diagnosis made? How long did the episode last? Were there other symptoms? Did he receive any treatment? Was a chest radiograph obtained? In fact, when I reviewed the radiology report for this case with Dr. Muse, it became apparent that a radiograph had been obtained as part of the evaluation for erythema nodosum and showed a right hilar fullness and a right paratracheal density, features possibly consistent with lymphadenopathy. The potential presence of lymphadenopathy at least 3 years earlier was reassuring and led me to feel more certain about the diagnosis in this case.

Erythema nodosum has been associated with bacterial infections, viral infections, mycoses, drug use (e.g., antibiotics and oral contraceptives), malignant diseases (e.g., lymphoma, leukemia, and other cancers), and other systemic diseases (e.g., sarcoidosis, enteropathies, Sweet's syndrome, and Takayasu's disease). Many diseases are associated with erythema nodosum, and among patients in whom a cause of erythema nodosum is identified, streptococcal infection and sarcoidosis are fairly common diagnoses. However, in most patients, an underlying diagnosis is not made. On the basis of this patient's history, streptococcal infection seems unlikely. However, the presence of lymphadenopathy and the history of erythema nodosum may point toward a diagnosis of sarcoidosis.

Sarcoidosis and Löfgren's Syndrome

When erythema nodosum developed in this patient, he also had ankle pain and, within a month, an episode of arthritis of the right knee that prompted joint aspiration. The combination of erythema nodosum and ankle pain should raise the possibility of a diagnosis of Löfgren's syndrome, although the presence of hilar lymphadenopathy is generally also required for this diagnosis. Löfgren's syndrome is a form of sarcoidosis that is characterized by the triad of erythema nodosum, arthralgias, and hilar lymphadenopathy; it is usually self-limited, although its manifestations may persist for several years. A small number of patients with Löfgren's syndrome have normal chest radiographs at the time of diagnosis; most patients have hilar lymphadenopathy, 13% of patients have respiratory symptoms, and 4% of patients have peripheral lymphadenopathyFurthermore, 6% of patients have a recurrence of sarcoidosis within 18 months to 20 years after presenting with Löfgren's syndrome. I think the episode of erythema nodosum, the ankle pain, and the knee arthritis were manifestations of Löfgren's syndrome, in which case, sarcoidosis would be the leading potential diagnosis.

Does sarcoidosis account for the other features of this patient's presentation ? Gastrointestinal manifestations of sarcoidosis primarily involve the stomach, and patients can present with abdominal pain, nausea, vomiting, and weight lossHeadaches, such as the one this patient reportedly had during the weeks before admission, are sometimes a manifestation of sarcoidosisThe patient's hypoattenuating splenic lesions are consistent with sarcoidosis. 

Are the findings on this patient's chest imaging consistent with sarcoidosis? Approximately 85% of patients with sarcoidosis have intrathoracic lymphadenopathy, typically bilateral hilar and right paratracheal lymphadenopathy. When I reviewed the radiograph before this conference, I was impressed by the predominance of the findings in the upper lung and the coexistent mediastinal lymphadenopathy, features that suggest sarcoidosis. Given the findings on the chest radiograph in this case, I strongly suspect that the diagnosis is sarcoidosis, although there are still a few issues to address.

We still need to consider the minimally elevated blood level of ACE. When the ACE level is less than 25 U per liter, the likelihood ratio for sarcoidosis is 0.12; when the ACE level is between 25 and 71 U per liter, the likelihood ratio is 1.31; and when the ACE level is greater than 71 U per liter, the likelihood ratio increases to 7.15. This patient had an ACE level of 55 U per liter, and therefore the probability of sarcoidosis was slightly increased, although the clinical and radiographic findings already made sarcoidosis an overwhelmingly likely diagnosis.

The next issue is to rule out the potential diagnoses of lymphoma and infection. The three possible areas to target for a tissue biopsy are the inguinal region, the chest (which could be evaluated with a transbronchial biopsy), and the spleen (which could potentially be removed). The safest area in which to perform a biopsy should be selected. The inguinal nodes would be the safest to access, although a transbronchial biopsy would also be a reasonable approach for obtaining a tissue sample in this case.

Dr. Eric S. Rosenberg (Pathology): Dr. Strom, what was your impression when you initially evaluated this patient?

Dr. Strom: We thought that this patient's abdominal pain was probably due to indigestion and was unrelated to the subacute disease process. Our differential diagnosis for the subacute disease process was sarcoidosis, tuberculosis, endemic fungal infection, granulomatous polyangiitis, and lymphoma. Given the presence of several enlarged lymph nodes and the development of night sweats, we were most concerned about lymphoma.

The history of fatigue may indicate an underlying systemic illness that preceded the episodes of acute pain, but it is nonspecific. The fact that there were two discrete episodes in which the patient had similar symptoms in different distributions suggests an underlying process that is being manifested in different anatomical locations over time. It is also noteworthy that the episodes involved pain in both the right and left lower quadrants of the abdomen and in the flank; pathologic conditions affecting the colon (e.g., mesenteric ischemia or inflammatory bowel disease) or the ureters or kidneys (e.g., nephrolithiasis or pyelonephritis) should be considered. Metabolic causes of abdominal pain, such as poisoning from heavy metals and the neurovisceral porphyrias, are also possibilities.

Alcohol is classically associated with pancreatitis, but the clinical presentation in this case is atypical for pancreatitis. The patient's extensive smoking history increases his risk of vascular disease, which can be manifested as mesenteric ischemia or infarction, but this scenario is more common in older patients. Although he reports no personal history of cardiovascular disease, an arrhythmia such as atrial fibrillation, which can be familial, could lead to embolic infarction in multiple vascular territories.

The patient's vital signs are reassuring, suggesting the absence of an immediately life-threatening condition, although he has marked hypertension. The hypertension could be a result of the pain or of the underlying abdominal process; if it does not improve with pain control, then antihypertensive medication should be initiated. The abdominal examination is also reassuring in that there are no signs of localized or general peritoneal inflammation (i.e., rebound tenderness or guarding), which would suggest organ perforation or abdominal infection. The bilateral tenderness at the costovertebral angles raises the possibility of retroperitoneal disease and could be due to bilateral renal disease (e.g., pyelonephritis or hydronephrosis), pancreatitis, nephrolithiasis, or a retroperitoneal hematoma, mass, or abscess.

The mild leukocytosis is nonspecific and could be a consequence of inflammation, infection, or pain. The normal liver-function tests and amylase and lipase levels make hepatobiliary and pancreatic disease unlikely, and the normal lactate level is helpful in ruling out bowel ischemia. The creatinine level of 1.5 mg per deciliter in an otherwise healthy 41-year-old man with no known medical problems suggests clinically significant renal dysfunction. The renal impairment is probably due to intrinsic renal disease or obstruction, given the low ratio of blood urea nitrogen to creatinine, and it could explain the patient's hypertension, which can be observed with renal injury or renal vascular insufficiency. The absence of hematuria makes nephrolithiasis unlikely, and the bland urinary sediment is inconsistent with glomerulonephritis. The elevated level of lactate dehydrogenase, a marker of cell turnover that is often elevated in patients with neoplasms, could indicate a retroperitoneal cancer, such as a lymphoma causing compressive hydronephrosis, but lactate dehydrogenase levels can also be elevated in patients with necrosis of the bowel or kidney. Abdominal imaging should be performed to determine whether the patient has any retroperitoneal or renal abnormalities.

The apparent scarring in the left kidney suggests that infarction of the left kidney preceded that of the right kidney, information that correlates with the two discrete episodes of abdominal pain. Hypertension is common in patients with acute renal infarction and is thought to result from increased release of renin, which suggests that the hypertension will respond well to treatment with angiotensin-converting–enzyme (ACE) inhibitors.

Although renal infarction can occur after isolated renal-artery thrombosis in patients with vascular disease, bilateral renal infarctions are suggestive of an embolic phenomenon, a primary pathologic process of the vasculature (e.g., a vasculitis or fibromuscular dysplasia), or a prothrombotic state. Echocardiography, electrocardiography, and continuous cardiac telemetry should be performed to determine whether the patient has a cardiac thrombus, valvular vegetations, or cardiac arrhythmias. A workup for a hypercoagulable state is also appropriate, as is an assessment for inflammatory markers. Given the risk of hemorrhagic stroke in patients with endocarditis, treatment with anticoagulant drugs should be deferred until the results of echocardiographic imaging have been obtained. Performance of echocardiography should be expedited.

A transthoracic echocardiogram was unremarkable. Treatment with unfractionated heparin was started. The patient's blood pressure improved after the initiation of treatment with an ACE inhibitor. Electrocardiography revealed a normal sinus rhythm. Blood and urine cultures were sterile. A panel of tests for prothrombotic activity was ordered, including tests of anticardiolipin IgG and IgM antibody levels, protein C activity, protein S antigen level, cryoglobulins, antithrombin 3 function, and lupus anticoagulant, as well as genetic testing for factor V Leiden. The erythrocyte sedimentation rate was 31 mm per hour (normal range, 0 to 12), and the C-reactive protein level was 193 mg per deciliter (normal value, <3).

In light of these results, the most likely cause of the patient's bilateral renal infarcts is a primary vascular process; nonetheless, anticoagulation therapy should be continued pending the results of the hypercoagulability studies. Although fibromuscular dysplasia remains a possibility, this condition typically affects young women. The patient's elevated C-reactive protein level and erythrocyte sedimentation rate suggest systemic inflammation, which would be consistent with a vasculitis. The primary vasculitides that typically affect the kidney include granulomatosis with polyangiitis (formerly known as Wegener's polyangiitis), polyarteritis nodosa, Henoch–Schönlein purpura (also known as IgA-associated vasculitis), microscopic polyangiitis, and cryoglobulinemic vasculitis. The absence of an inflammatory urinary sediment argues against vasculitides with arteriolar or capillary involvement (e.g., granulomatosis with polyangiitis), since in these conditions renal involvement is manifested as glomerulonephritis. An initial laboratory workup for patients in whom vasculitis is suspected should include serologic tests for indicators that are elevated in the presences of vasculitides and for infections associated with vasculitis. Cocaine and certain medications, such as ergot derivatives, can lead to a presentation that mimics vasculitis; the patient should be questioned again about any use of illicit drugs or medications, including herbal supplements, and a drug screen should be obtained. He should also be examined for evidence of vasculitic organ involvement, such as skin lesions or mononeuritis multiplex. If such findings are present, a nerve, skin, or muscle biopsy can be performed. Since a risk of hemorrhage or a potentially catastrophic vascular rupture is associated with renal and hepatic biopsies in patients with polyarteritis nodosa, visceral angiography should be considered if there is no evidence of involvement of skin, nerve, or muscle tissue amenable to biopsy.

The results of screens for antinuclear antibodies and antineutrophil cytoplasmic antibodies (ANCAs), hepatitis B virus (core antibody, surface antibody, and surface antigen), hepatitis C virus, and the human immunodeficiency virus were unremarkable, as were the results of the hypercoagulability panel. Levels of complement (C3 and C4) were normal. A 24-hour urine collection revealed 232 mg of protein. A urine toxicology screen was negative. Thorough physical examination revealed no skin lesions or neurologic deficits. Abdominal angiography revealed numerous microaneurysms in multiple vascular territories, including microaneurysms in both kidneys and in the hepatic and superior mesenteric arteries; these findings are consistent with a diagnosis of polyarteritis nodosa (Figure 2FIGURE 2Angiographic Studies of the Vasculature of the Left Kidney and the Superior Mesenteric Artery.). Anticoagulation therapy was discontinued. Treatment with intravenous methylprednisolone and oral cyclophosphamide was initiated; creatinine levels showed a downward trend from the peak value of 1.5 mg per deciliter, with complete resolution of abdominal pain.

The five-factor score (with a point each for the presence of proteinuria at >1 g of protein per day, creatinine level >1.58 mg per deciliter [139.7 μmol per liter], cardiomyopathy, severe gastrointestinal involvement [e.g., bleeding or infarction], and central nervous system involvement) is a prognostic tool that can be used to inform treatment decisions for a number of vasculitides, including polyarteritis nodosa. Single-agent (glucocorticoid) therapy is often used as the initial treatment strategy in patients with a five-factor score of 0, although many of these patients will require an additional agent. This patient might be considered to have a score of 0, but the mesenteric involvement arguably raises his score to 1, and the angiographic findings of renal and mesenteric involvement raise the possibility of a potentially catastrophic gastrointestinal complication, given that visceral involvement is the predominant factor in predicting prognosis. Thus, the decision on the part of the treating physician to use aggressive immunosuppressive therapy appears to be reasonable.

The patient was discharged with prescriptions for oral prednisone and cyclophosphamide, with plans for close follow-up. When he returned for an assessment 4 months later, the C-reactive protein level and erythrocyte sedimentation rate had normalized, and the creatinine level remained stable. Treatment with cyclophosphamide and prednisone continued for another 5 months, at which point a transition was made to treatment with mycophenolic acid and prednisone. At follow-up 10 months after diagnosis, treatment with mycophenolic acid was temporarily withheld because of a perirectal abscess. With prednisone alone, the patient had recurrent abdominal pain, which responded to the resumption of treatment with mycophenolic acid. At follow-up 14 months after diagnosis, the patient was taking mycophenolic acid and prednisone and was free of symptoms.


Given the protean manifestations of the vasculitides, they should be considered in the differential diagnosis of a range of clinical presentations. In this patient, the finding of infarcts in both kidneys, along with laboratory evidence of marked inflammation, led to renal angiography, which showed evidence of numerous microaneurysms characteristic of polyarteritis nodosa.

The vasculitides are characterized by inflammation of the blood-vessel wall, leading to vessel damage, eventual luminal compromise, and subsequent downstream ischemia; their specific clinical manifestations depend on the size, type, and location of the affected blood vessels.1 The classification of the vasculitides is based on the size and type of the affected vessels and on the presence or absence of ANCAs.2 Takayasu's arteritis, for example, is an ANCA-negative, large-vessel vasculitis primarily affecting the aorta and its principal branches, whereas granulomatosis with polyangiitis is an ANCA-associated vasculitis that predominantly affects small arteries of the lungs and kidneys.

Polyarteritis nodosa is an ANCA-negative vasculitis that typically affects medium-size arteries and may affect any organ system. As is the case with other vasculitides, patients with polyarteritis nodosa typically have systemic symptoms (e.g., fatigue, arthralgias, weight loss, or fever) and may also have neurologic symptoms (peripheral neuropathy or mononeuritis multiplex), skin involvement (nodules [often necrotizing], purpura, or livedo), renal involvement, or a combination of these symptoms.3 One of the most commonly affected organs is the kidney. There are no formal diagnostic criteria for polyarteritis nodosa.The diagnosis is made on the basis of a constellation of typical symptoms, supportive laboratory tests, and either a biopsy of the affected tissue (i.e., skin, nerve, or muscle tissue — both renal and hepatic biopsies are contraindicated) showing characteristic necrotizing arterial inflammation or visceral angiography showing microaneurysms in the renal, mesenteric, or hepatic vasculature. These aneurysms can be difficult to detect, since they may be small, limited to a single organ or vascular territory, and few in number. 4 Laboratory tests are supportive but not diagnostic and may suggest alternative diagnoses. Inflammatory markers (e.g., the erythrocyte sedimentation rate and the level of C-reactive protein) are often elevated but are nonspecific.5 The presence of ANCAs against myeloperoxidase or against protease 3 favors the diagnosis of microscopic polyangiitis and granulomatosis with polyangiitis, respectively, whereas low complement levels suggest cryoglobulinemic vasculitis (low C4 levels) or systemic lupus erythematosus (low levels of C3 and C4).

As is true of other vasculitides, the pathogenesis of polyarteritis nodosa is poorly understood. Some cases of polyarteritis nodosa are associated with hepatitis B, but most are idiopathic.

Data from randomized trials and observational studies suggest that treatment with glucocorticoids alone or in combination with cyclophosphamide or azathioprine induces remission in the majority of patients with polyarteritis nodosa. 6,7 The specific treatment generally depends on the severity of the disease. Although five-factor scores were designed to assess the severity of disease and the prognosis, clinicians have adapted the criteria to guide therapy, despite a lack of robust prospective data supporting this approach. In the one randomized treatment trial, glucocorticoid monotherapy was sufficient to induce remission in 50% of patients with mild polyarteritis nodosa (five-factor score of 0), but approximately 40% required additional immunosuppression because of relapse.7Combination therapy is used in patients with more severe manifestations8 and in those with serious findings that are not reflected in the five-factor score (in this case, renal infarction). It remains unclear, however, which of the factors that are not included in the five-factor score should influence treatment decisions and how treatment should be modified to accommodate such factors, although visceral involvement does appear to warrant more aggressive therapy.9 In the absence of comparative trials to guide treatment duration, typical practice involves several weeks of high-dose prednisone followed by tapering over a period of 6 months to a year; treatment with cyclophosphamide is generally limited to 1 year, given concerns about toxicity; other, less toxic agents may be used in conjunction with prednisone once cyclophosphamide has been discontinued. Hypertension is a common manifestation of polyarteritis nodosa that can be effectively treated with ACE inhibitors, as was done in this patient.10

The prognosis for patients with untreated polyarteritis nodosa is grim, with a 5-year survival rate of 13%; death is often a consequence of renal failure, myocardial infarction, or stroke.11,12 With appropriate therapy, the 5-year survival rate is approximately 80%. 13 Given the poor prognosis associated with untreated polyarteritis nodosa and the substantial improvement with appropriate treatment, it is important to consider this condition early in the differential diagnosis in order to minimize morbidity and the risk of death. In this case, the prompt recognition of atypical features of the patient's presentation with abdominal pain (i.e., pain occurring in different distributions over time) suggested that vasculitis was a diagnostic possibility and prompted imaging of the mesenteric and renal vasculature, which led to the diagnosis and administration of effective therapy.