|Home | About | Journals | Submit | Contact Us | Français|
A 66-year-old woman presented for evaluation of jaundice that had been present for 6 weeks. She first noticed that her urine began to appear darker and then experienced pruritus under her breasts that gradually spread throughout her body. She sought evaluation at an urgent care center and was initially thought to have scabies. Two weeks later, she noted a yellowing of her skin and began having gray-colored stools. Also at this time, she had mild right upper quadrant abdominal discomfort, nausea, and fullness associated with eating fatty food. She modified her diet to mainly carbohydrates and noted resolution of the gastrointestinal symptoms.
The patient had no history of hepatitis B or C, alcohol use, fevers, chills, anorexia, vomiting, myalgias, sick contacts, travel outside the United States, or acetaminophen usage. There was no family history of gastrointestinal or autoimmune disease. Her medical history was notable for idiopathic focal sclerosing glomerulonephritis that was diagnosed with renal biopsy when proteinuria was discovered incidentally. Lisinopril therapy was prescribed. At that same time, an antinuclear antibody (ANA) test yielded weakly positive results. She reported having a “liver problem” in the distant past that was associated with an oral antifungal medication and lasted 8 to 9 months, but she could not recall further details.
Vital signs were normal on admission. Physical examination revealed an older woman who was not in acute distress but had scleral icterus, skin that was diffusely jaundiced, most prominent on the chest and back, and many excoriations on her back, arms, and legs. Her abdomen was soft, nondistended, and mildly tender to palpation of the right upper quadrant with no rebound or guarding. Normoactive bowel sounds and palpable hepatomegaly 4 to 5 cm below the costal margin were noted. There was no splenomegaly.
The differential diagnosis for jaundice can be described by all of these etiologies. Infections with bacteria and parasites may cause jaundice. For instance, Weil syndrome, an infection caused by Leptospira, a spirochete, may present with jaundice, myalgias, fevers, myocarditis, and rash.1 This would not be a likely cause in this patient because she does not have these symptoms. Moreover, other infections such as acute cholecystitis or cholangitis are also unlikely. Acute cholecystitis may present with painful jaundice when there is gallstone obstruction of the cystic duct. If the gallstone is large and is in a Hartmann pouch (an out-pouching of the wall of the gallbladder at the junction of the neck of the cystic duct), the stone can cause external compression of the common hepatic duct; this is commonly known as Mirizzi syndrome, a complication of cholelithiasis.2 This is unlikely in this patient given that her acute abdominal pain quickly resolved although her jaundice persisted and she had hepatomegaly on examination, which is not a finding associated with cholecystitis. Cholangitis, caused by infection and obstruction of the common bile duct, presents with the Charcot triad of fever, jaundice, and right upper quadrant pain.3 In general, patients with cholangitis appear quite ill. Our patient did not have fevers and was not ill-appearing on examination, making cholangitis unlikely.
Pancreatic cancer or a bile duct malignancy may present with jaundice as the tumor grows and obstructs the common bile duct. Other associated symptoms are unintentional weight loss and constitutional symptoms. Although this patient did not have weight loss, given her age and presentation of painless jaundice, malignancy is high on the differential diagnosis and is the most likely cause. Autoimmune diseases may also result in painless jaundice; however, given the patient's age, no family history of autoimmune disease, and lack of laboratory results at this time, an autoimmune etiology is less likely.
Although metabolic etiologies are possible, including Wilson disease and Gilbert syndrome, these are both unlikely in a woman of this age. In addition, Wilson disease may be accompanied by Kayser-Fleischer rings, which were not observed on physical examination. Gilbert syndrome, caused by decreased activity of the enzyme UDP-glucuronyl transferase, results in an unconjugated hyperbilirubinemia.3 It presents as mild painless jaundice when there is acute stress on the body.3 This is unlikely in our patient because she presented with a change in her stool and urine color, which is consistent with a conjugated bilirubinemia.3 Toxins such as alcohol and acetaminophen overdose may certainly result in jaundice; however, this patient denies use of both.
Initial laboratory studies yielded the following results (reference ranges provided parenthetically): white blood cell count, 5.1 × 109/L (3.5-10.5 × 109/L); hemoglobin, 11.7 g/dL (12.0-15.5 g/dL); creatinine, 1.5 mg/dL (baseline, 1.8 mg/dL) (0.6-1.1 mg/dL); serum urea nitrogen, 30 mg/dL (6-21 mg/dL); international normalized ratio, 1.1 (0.8-1.2); total bilirubin, 22.2 mg/dL (0.1-1.0 mg/dL); direct bilirubin, 14.8 mg/dL (0.0-0.3 mg/dL); aspartate aminotransferase (AST), >700 U/L (8-43 U/L); amylase, 123 U/L (26-102 U/L); lipase, 59 U/L (10-73 U/L); and alkaline phosphatase, 358 U/L (55-142 U/L). Hepatitis serologies revealed the following: hepatitis A IgM antibody, negative; hepatitis B core IgM and total antibody, negative; hepatitis B surface antibody, positive; hepatitis B surface antigen, negative; and hepatitis C antibody screen, negative.
Ultrasonography of the abdomen showed a normal gallbladder with intraluminal debris. The common bile duct was mildly to moderately dilated, measuring 1.0 cm in maximal luminal diameter. Findings on computed tomography (CT) of the abdomen were unremarkable. Because work-up results were unrevealing, we decided to proceed with endoscopic retrograde cholangiopancreatography (ERCP) to better evaluate the common bile and pancreatic ducts, which could be stented if there was an obstruction. However, ERCP findings were unremarkable.
Biopsy of the liver would not be recommended at this time because it is invasive and carries associated risks. Since no pancreatic lesions were noted on CT and no obstructing lesion was noted on ERCP, pancreatic carcinoma is low on the differential diagnosis, and thus pancreatic biopsy would have low yield. In addition, biopsy of the pancreas increases the patient's risk of procedure-related complications, including severe pancreatitis. While noninvasive, MRCP would not provide any additional information because ERCP was already performed. The presence of antimitochondrial antibodies would be consistent with primary biliary cirrhosis that characteristically presents with pruritus and jaundice in women of middle age.4 Therefore, investigation of autoimmune causes of cholestatic hepatitis would be warranted.
If there is high suspicion for a structural lesion that would benefit from intervention such as stenting, ERCP should be performed. Endoscopic retrograde cholangiopancreatography should not be considered a diagnostic test but rather a therapeutic tool. If only cholangiography is desired or there is a low pretest probability of finding a structural lesion amenable to endoscopic treatment, then MRCP or endoscopic ultrasonography is the appropriate diagnostic modality rather than ERCP. However, before these tests are done, serologic tests are usually undertaken to help guide clinical decision making regarding the usefulness of further testing and procedures.
A number of parasites can cause biliary obstruction. For example, Clonorchis sinensis, also known as the Oriental liver fluke, infects humans via the consumption of infected raw fish and causes symptoms as a result of bile duct obstruction by the adult fluke.5 Testing stool for ova and parasites would not be of high yield in this patient since she did not consume raw fish and was not from an endemic area.
Further laboratory analysis showed a positive anti–smooth muscle antibody titer (1:160), positive ANA test result (2.9 U) (≤1.0 U), negative antimitochondrial antibody titer, and increased IgG level (>2000 mg/dL). Testing was positive for Epstein-Barr virus IgM antibody and negative for Epstein-Barr viral load.
The diagnosis of autoimmune hepatitis is supported by the patient's clinical presentation, hyperbilirubinemia, positive anti–smooth muscle antibody titer, elevated ANA concentration, and increased IgG level.6-9 Primary sclerosing cholangitis may present with similar clinical symptoms but is more commonly seen in men with inflammatory bowel disease, especially ulcerative colitis, and many patients have antineutrophilic cytoplasmic antibodies3; however, ERCP did not reveal evidence of ductal changes to suggest primary sclerosing cholangitis.3,4 Herpes simplex hepatitis, a rare disease associated with considerable mortality in immunosuppressed patients, can present with nonspecific findings, elevated transaminase concentrations, low bilirubin levels, and hypodensities noted on CT imaging.10,11 Primary biliary cirrhosis is an autoimmune disease seen more commonly in women and is associated with a positive antimitochondrial antibody titer. While it is possible to develop jaundice with primary biliary cirrhosis, pruritus will generally precede the jaundice.3
Once there is high clinical suspicion for autoimmune hepatitis, a liver biopsy is performed before treatment initiation to obtain a baseline evaluation of hepatic involvement. Therefore, our patient underwent a liver biopsy that showed a grade 3/4 moderate, active, chronic hepatitis with stage 2/4 fibrosis. At presentation, her ferritin level was markedly elevated at 1231 μg/L (11-307 μg/L), representing an inflammatory reaction; there was no hepatic excess of iron.
When initiating treatment, it is important to evaluate the severity of the patient's disease with goals in mind. With active disease, as seen in our patient, the primary goal is to achieve remission. Infliximab is a monoclonal antibody against tumor necrosis factor α used in the treatment of many autoimmune conditions; however, it is not a first-line therapy for autoimmune hepatitis. Corticosteroids such as prednisone are commonly used to achieve remission and are the first line of treatment at doses up to 60 mg/d for 4 weeks followed by slow tapering.7 Azathioprine, an immunosuppressant that is metabolized to 6-mercaptopurine, is primarily used as bridging therapy for patients being transitioned off corticosteroids.7 Hydroxyzine, an antihistamine, may be used in the symptomatic control of pruritus but will not impact disease progression. Liver transplant should be reserved for rare presentations with fulminant hepatic failure unresponsive to therapy and for end-stage liver disease. In our patient, prednisone therapy at 60 mg/d yielded dramatic results. At discharge, her total bilirubin level was 8.7 mg/dL, direct bilirubin was 4.8 mg/dL, AST was 275 U/L, and alanine aminotransferase (ALT) was 580 U/L. She was also vaccinated against hepatitis A.
Management of autoimmune hepatitis must be clinically tailored to the patient. Long-term corticosteroids cause unwanted adverse effects. Ideally, prednisone could be prescribed initially and then transitioned to another therapy. The use of infliximab for maintenance therapy requires further investigation.6 Tapering corticosteroids and initiating azathioprine is an excellent choice because azathioprine is a steroid-sparing agent and will help maintain remission while the corticosteroid is tapered, as was done in this patient.6 Azathioprine can be started at 1 to 2 mg/kg per day.6 Thiopurine methyltransferase is an enzyme that metabolizes azathioprine. Its level must be checked before the initiation of azathioprine because of the risk of bone marrow toxicity.7 Tacrolimus is generally used in patients in whom first-line therapies have failed.7 Therapeutic plasma exchange involves removing larger molecules such as antibodies from the blood in order to prevent further damage and is generally used as a last resort.12 Our patient was started on a regimen of azathioprine 50 mg daily before hospital dismissal.
This case illustrates the importance of the steps a clinician must take in developing a differential diagnosis for painless jaundice. Our patient underwent ERCP, which was not necessary to diagnose the autoimmune hepatitis. In retrospect, after CT and ultrasound imaging failed to demonstrate a structural lesion, testing serologies for autoimmune diseases should have been performed. If the diagnosis still remained in question, then a diagnostic MRCP could have been ordered.
The initial evaluation of painless jaundice involves obtaining a complete blood cell count, measuring levels of serum AST, ALT, total and direct bilirubin, γ-glutamyl transpeptidase, alkaline phosphatase, and performing urinalysis to help decide if this is a cholestatic (obstructive) process or not.3 Only conjugated bilirubin is able to pass into the urine; therefore, in a patient with an elevated total bilirubin level, normal direct bilirubin concentration, and negative results on urinalysis, one must investigate causes of an unconjugated hyperbilirubinemia such as hemolysis.3
Conversely, if the total bilirubin and direct bilirubin levels are elevated, conjugated hyperbilirubinemia should be considered.3 Elevation of γ-glutamyl transpeptidase and alkaline phosphatase concentrations provides clues for an obstructive process.3 Once an obstructive process is suspected, noninvasive imaging should be considered. Ultrasonography should be performed first because it is more cost-effective, noninvasive, and the most sensitive imaging technique for detection of stones, whereas CT with contrast agent is better for detecting parenchymal disease of the liver.3 Neither have sufficient sensitivity to rule out an obstructing intraductal stone or mass, and therefore, ERCP or MRCP may be required.3 If there is a low pretest probability of finding a structural lesion that may be amenable to intervention, then one should proceed with MRCP.
In patients who present with laboratory findings suggestive of an obstructive process but no structural lesion on imaging, hepatitis should be considered. Transaminases may provide clues to the etiology.3 Transaminase levels that are over 1000 U/L are suspicious for viral hepatitis, ischemic hepatitis, and toxin-induced liver injury, and levels greater than 10,000 U/L raise suspicion for hepatic failure.3
Fulminant hepatic failure that may be a result of shock or acetaminophen overdose is suspected when the transaminase level is greater than 10,000 U/L in combination with symptoms suggestive of synthetic hepatic failure such as jaundice and encephalopathy.3 Other etiologies of acute liver failure include hepatitis A and B viral infection and Reye syndrome, which is generally associated with aspirin use in younger individuals. For at-risk individuals, hepatitis screening should be performed as well as measurement of serum ferritin and ceruloplasmin levels. If the etiology remains unclear, testing for autoantibodies such as ANA, anti–smooth muscle antibody, and antimitochondrial antibodies is recommended, as was done in this patient.3 If the serology results suggest an autoimmune disease such as autoimmune hepatitis, a liver biopsy should be done to evaluate baseline hepatic involvement before treatment initiation.
In general, in patients with autoimmune hepatitis, liver function test results may fluctuate over a long period of time; therefore, this patient's remote history of a “liver problem” may have been the initial start of the autoimmune hepatitis. Moreover, because of disease chronicity, many patients will have fibrosis evident on biopsy.
See end of article for correct answers to questions.
CORRECT ANSWERS: 1. b. 2. d. 3. a. 4. b. 5. c