SYMPTOMS TO DIAGNOSIS

GREGORY W. RUTECKI, MD, Section Editor

WILLIAM C. LIPPERT, MD, MPH EUN Y. LEE, MD AIBEK E. MIRRAKHIMOV, MD
Instructor of Internal Medicine, Department Professor and Director of Anatomic Assistant Professor of Medicine, Division of
of Internal Medicine, Section of Hospital Pathology, Department of Pathology Hospital Medicine, University of Kentucky
Medicine, Wake Forest Baptist Medical Center, and Laboratory Medicine, University of Medical Center, Lexington, KY
Winston-Salem, NC Kentucky Medical Center, Lexington, KY

A woman, age 35,

with new-onset ascites
35-year-old woman is admitted to the noted mild elevation of AST and ALT, as well
A hospital with a 5-day history of abdomi-
nal distention and jaundice. She reports no
as an elevated international normalized ratio
(INR) and hyperbilirubinemia. Computed to-
history of fever, chills, night sweats, abdomi- mography of the abdomen and pelvis showed
nal pain, nausea, vomiting, diarrhea, changes hepatomegaly with possible fatty liver. Because
in urine color, change in stool color, weight of these results, the patient was transferred to
loss, weight gain, or loss of appetite. our institution for further evaluation.
She is petite, with a body mass index of
19.4 kg/m2. She has no known history of med- ■ EVALUATION AT OUR INSTITUTION
ical conditions or surgery and is not taking any On examination at our institution, she is afe-
medications. Her family history is unremark- brile, and vital signs are within normal ranges.
able, and she denies current or past tobacco, She has bilateral scleral icterus and diffuse
alcohol, or illicit drug use. jaundice, but no other skin finding such as
rash or spider angioma. She has no lymph-
■ RECENT TRAVEL adenopathy. Her abdomen is distended, with A serum-ascites
She says that during a trip to Central America tense ascites, and her liver is tender to palpa- albumin
several months ago, she had suffered a seizure tion. The tip of the spleen is not palpable.
and was taken to a local hospital, where labo- The cardiovascular examination reveals gradient
ratory testing revealed elevated aspartate ami- no murmurs, rubs, or gallops, but she has jugu- ≥ 1.1 g/dL
notransferase (AST) and alanine aminotrans- lar venous distention and +2 pitting edema of
ferase (ALT) levels. She says that the rest of both lower extremities. indicates portal
the workup at that time was normal. On respiratory examination, there is dull- hypertension
About 1 week after that incident, she re- ness to percussion, with slight crackles on aus-
turned home and saw her primary care phy- cultation at the right lung base. The neuro-
sician, who ordered further testing, which logic examination is normal.
showed mild hyperbilirubinemia and mild Table 1 shows the results of initial labora-
elevation of AST and ALT levels. Her physi- tory testing.
cian attributed the elevations to atovaquone,
which she had been taking for malaria prophy-
laxis, as repeat testing 2 weeks later showed
improvement in AST and ALT levels.
1 Which study would provide the most in-
formation on the cause of ascites?
□ Abdominal ultrasonography
The patient says she returned to her normal □ Abdominal paracentesis with ascitic fluid
state of health until about 5 days ago, when analysis
she noticed jaundice and abdominal disten- □ Chest radiography
tion, but without abdominal pain, dark urine, □ Echocardiography
or clay-colored stools. She became concerned □ Urine protein-to-creatinine ratio
and went to her local hospital. Testing there
Abdominal paracentesis with ascitic fluid
doi:10.3949/ccjm.86a.17082 analysis is the essential study for any patient

CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 86 • NUMBER 4 APRIL 2019 257

Downloaded from www.ccjm.org on April 26, 2022. For personal use only. All other uses require permission.
 NEW-ONSET ASCITES

TABLE 1 normal kidney sizes, no hydronephrosis, and

no intra-abdominal mass. Chest radiography
Results of initial laboratory testing is clear with no sign of consolidation, edema,
or effusion. Echocardiography shows a normal
Reference left ventricular ejection fraction with no val-
Test Resulta range
vular disease or pericardial effusion. A random
White blood cell count 17.3 x 109/L (3.7–10.3) urine protein-creatinine ratio is normal at 0.1
Neutrophil count 82% (reference range < 0.2).
Hemoglobin 8.8 g/dL (11.2–15.7)
Hematocrit
Platelet count
Sodium
24.0%
112 x 109/L
128 mmol/L
(34%–45%)
(155–369)
(136–145)
2 What is the most likely cause of her ascites
based on the workup to this point?
□ Cirrhosis
Potassium 3.2 mmol/L (3.7–4.8)
□ Heart failure
Chloride 86 mg/dL (101–108)
Blood urea nitrogen 16 mg/dL (7–21) □ Nephrotic syndrome
Creatinine 0.62 mg/dL (0.60–1.10) □ Portal vein thrombus
Glucose 91 mg/dL (74–99) □ Abdominal malignancy
Aspartate aminotransferase 97 U/L (11–32) □ Malaria
Alanine aminotransferase 24 U/L (8–33)
Alkaline phosphatase 224 U/L (35–104) An initial approach to ascitic fluid analysis is
Gamma-glutamyltransferase 996 U/L (7–36) to calculate the serum-ascites albumin gradient
Total bilirubin 17.0 mg/dL (0.2–1.1) (SAAG). The SAAG is calculated as the serum
Total protein 4.9 g/dL (6.3–7.9) albumin level minus the ascitic fluid albumin
Albumin 2.2 g/dL (3.3–4.6) level.4,5 This is useful in determining the cause
Total calcium 7.3 mg/dL (8.9–10.2) of the ascites (Figure 1).4,5 A gradient of 1.1 g/
Erythrocyte sedimentation rate 16 mm/h (0–20) dL or higher indicates portal hypertension.4,5
C-reactive protein 4.2 mg/dL (0–0.9)
Common causes of portal hypertension in-
Brain natriuretic protein 445 pg/mL (0–449)
Alpha-1 antitrypsin 2.8 ng/mL (0.9–9.0) clude cirrhosis, alcoholic hepatitis, heart fail-
Ceruloplasmin 30 mg/dL (18–53) ure, vascular occlusion syndromes (eg, Budd-
Ferritin 302 ng/dL (13–150) Chiari syndrome, portal vein thrombosis),
Hepatitis panel (A, B, C) Negative idiopathic portal fibrosis, and metastatic liver
Prothrombin time 20.2 sec (12.3–14.8) disease.5,6
International normalized ratio 1.7 sec (0.9–1.1) If portal hypertension is present based on
Partial thromboplastin time 38 sec (25–36) the SAAG, the next step is to review the as-
Thyroid-stimulating hormone 3.9 μIU/mL (0.4–4.2) citic protein level to help distinguish between
a hepatic and a cardiac etiology of the asci-
a
Abnormal results in boldface type. tes. An ascitic protein level less than 2.5 g/
dL indicates a primary liver pathology (eg, cir-
with clinically apparent new-onset ascites.1–3 rhosis). An ascitic protein level of 2.5 g/dL or
It is the study that provides the most informa- greater typically indicates a cardiac condition
tion on the cause of ascites. (eg, heart failure, pericardial disease) with sec-
In our patient, abdominal paracentesis ondary congestive hepatopathy.5,6
yields 1,000 mL of straw-colored ascitic fluid, If the SAAG is less than 1.1 g/dL, the asci-
tes is likely not from portal hypertension. Typ-
and analysis shows 86 nucleated cells, 28 of
ical causes of a low SAAG include infection,
which are polymorphonuclear cells, and 0 red
malignancy, pancreatic ascites, and nephrotic
blood cells, with negative Gram stain and cul- syndrome.5,6
ture. The ascitic albumin level is 0.85 g/dL, In our patient, the SAAG is 1.35 g/dL (2.2
with an ascitic protein of 1.1 g/dL. g/dL minus 0.85 g/dL), ie, elevated and due to
Abdominal ultrasonography shows a dif- portal hypertension. With an SAAG of 1.1 g/
fusely echogenic liver, no focal lesions, mod- dL or greater and an ascitic fluid protein level
erate ascites, normal portal vein flow, no intra- less than 2.5 g/dL, as in our patient, the most
hepatic or extrahepatic biliary duct dilation, likely cause is cirrhosis.
258 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 86 • NUMBER 4 APRIL 2019

Downloaded from www.ccjm.org on April 26, 2022. For personal use only. All other uses require permission.
 LIPPERT AND COLLEAGUES

Calculate SAAG

Gradient ≥ 1.1 g/dL Gradient < 1.1 g/dL

indicates portal hypertension excludes portal hypertension

Ascitic protein Ascitic protein • Infection

≥ 2.5 g/dL < 2.5 g/dL • Malignancy
• Nephrotic syndrome
• Pancreatic ascites
• Heart failure • Cirrhosis
• Pericardial disease
• Veno-occlusive disease

Figure 1. Interpreting the serum-ascites albumin gradient (SAAG) and ascitic protein levels.

Heart failure is unlikely based on her nor- □ Any patient with cirrhosis and an
mal brain natriuretic peptide level, an ascitic ascitic fluid protein level below 2.0 g/dL
fluid protein level below 2.5 g/dL, and nor- □ Any patient with cirrhosis and a history
mal results on echocardiography. Nephrotic of SBP
syndrome is also very unlikely based on the
patient’s normal random urine protein-cre- Any patient with cirrhosis and a history of
atinine ratio. Portal vein thrombus and ab- SBP should receive prophylactic antibiotics,8 Transjugular
dominal malignancy are essentially ruled out as should any patient deemed at high risk of
by the negative results of Doppler abdominal SBP. It is indicated in the following patients: biopsy
ultrasonography, with normal venous flow and • Patients with cirrhosis and gastrointestinal confirmed
no intra-abdominal mass and coupled with an bleeding9,10 alcoholic hepa-
elevated SAAG. • Patients with cirrhosis and a previous epi-
Although the patient has a history of trav- sode of SBP8 titis, despite
el, the incubation period for malaria would • Patients with cirrhosis and an ascitic fluid the patient’s
not fit the time frame of presentation. Also, protein level less than 1.5 g/dL with either
she did not have typical malarial symptoms, impaired renal function (creatinine ≥ 1.2 initial denial
her rapid malaria test was negative, and a pe- mg/dL, blood urea nitrogen level ≥ 25 mg/ of alcohol use
ripheral blood smear for blood parasites was dL, or serum sodium ≤ 130 mmol/L) or
negative. It should be noted, however, that liver failure (Child-Pugh score ≥ 9 and a
Plasmodium malariae infection classically pres- bilirubin ≥ 3 mg/dL)9
ents with flulike symptoms and can resemble • Patients with cirrhosis who are hospital-
nephrotic syndrome, including peripheral ized for other reasons and have an ascitic
edema, ascites, heavy proteinuria, hypoalbu- protein level < 1.0 g/dL.9
minemia, and hyperlipidemia.7 Our patient has no signs or symptoms of
gastrointestinal bleeding and no history of SBP.
3 In which patients is antibiotic prophylaxis
against spontaneous bacterial peritonitis
(SBP) appropriate?
Her ascitic fluid protein level is 1.1 g/dL, and she
has normal renal function. However, her Child-
Pugh score is 12 (3 points for total bilirubin > 3
□ Any patient with cirrhosis mg/dL, 3 points for serum albumin < 2.8 g/dL, 2
□ Any patient with cirrhosis who is points for an INR 1.7 to 2.2, 3 points for moder-
hospitalized ate ascites, and 1 point for no encephalopathy),
CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 86 • NUMBER 4 APRIL 2019 259

Downloaded from www.ccjm.org on April 26, 2022. For personal use only. All other uses require permission.
 NEW-ONSET ASCITES

To confirm this diagnosis, she subsequently

undergoes transjugular liver biopsy, consid-
ered the gold standard for the diagnosis of al-
coholic hepatitis. During the procedure, the
hepatic venous pressure gradient is measured
at 18 mm Hg (reference range 1–5 mm Hg),
suggestive of portal hypertension. The pathol-
ogy study shows severe fatty change, active
steatohepatitis with ballooning degeneration,
easily identifiable Mallory-Denk bodies, and
prominent neutrophilic infiltration, as well as
extensive bridging fibrosis (Figure 2). These
findings point to alcoholic hepatitis.
After the biopsy results, we speak with the
patient further about her alcohol habits. At
this point, she informs us that she has con-
sumed significant amounts of alcohol since
the age of 18 (6 to 12 alcoholic beverages per
day, including beer and hard liquor). There-
fore, based on this new information, on her
jaundice and ascites, and on results of labora-
tory testing and biopsy, we confirmed our diag-
nosis of alcoholic hepatitis.

4 When is drug treatment appropriate for al-

coholic hepatitis?
□ Model for End-stage Liver Disease
(MELD) score greater than 12
□ MELD score greater than 15
□ Maddrey Discriminant Function score
greater than 25
□ Maddrey Discriminant Function score
greater than 32
□ Glasgow score greater than 5
□ Glasgow score greater than 7
The best answer is a Maddrey Discriminant
Figure 2. Transjugular biopsy study (top) shows severe
steatosis from lipid accumulation in hepatocytes (arrow) Function score greater than 32. A variety of
(hematoxylin and eosin, × 100). Higher magnification (bot- scoring systems have been used to assess the
tom) shows inflammatory cells (neutrophils, arrows) sur- severity of alcoholic hepatitis and to guide
rounding Mallory-Denk bodies, which are hyaline (eosino- treatment, including the Maddrey Discrimi-
philic) inclusions (hematoxylin and eosin, × 400). nant Function score, the MELD score, and
the Glasgow score.11–16 They share similar
with a bilirubin of 17.0 mg/dL. Based on this, laboratory values in their calculations, includ-
she is placed on antibiotic prophylaxis for SBP. ing prothrombin time (or INR) and total bili-
Our patient then undergoes an extensive rubin.11–16 Typically, a Maddrey Discriminant
workup for liver disease. Results of tests for Function score greater than 32, a Glasgow
toxins, autoimmune diseases, and inheritable score of greater than 9, or a MELD score
diseases are all within normal limits. At this greater than 21 is used to determine whether
point, despite the patient’s reported negative pharmacologic treatment is indicated.11–16
alcohol history, our leading diagnosis is alco- The typical treatment is prednisolone or
holic hepatitis. pentoxifylline.11,17–21 The Lille score is de-
260 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 86 • NUMBER 4 APRIL 2019

Downloaded from www.ccjm.org on April 26, 2022. For personal use only. All other uses require permission.
 LIPPERT AND COLLEAGUES

Clinical signs and symptoms

Acute onset of jaundice (< 3 months), fever, ascites, proximal muscle loss, hepatic encephalopathy,
enlarged and tender liver, neutrophilic leukocytosis, elevated aspartate aminotransferase-to-alanine aminotransferase
ratio (> 2:1), hyperbilirubinemia, elevated international normalized ratio with a history of alcohol use

Assess severity
Use the Maddrey Discriminant Function (DF) score, Model for End-Stage
Liver Disease (MELD) score, or Glasgow score

DF score ≥ 32, MELD score ≥ 21, Glasgow score ≥ 9 DF score < 32, MELD score < 21, Glasgow score < 9

Consider treatment Supportive care

Prednisolone 40 mg/day
Or
Pentoxyfylline 400 mg 3 times per day

Assess response to prednisolone

Calculate the Lille score at 1 week

Lille score ≤ 0.45 Lille score > 0.45

Continue prednisolone Stop prednisolone

Total duration of 28 days

Figure 3. Algorithm for the management of alcoholic hepatitis.

signed to help decide whether to stop cortico- ■ ALCOHOLIC HEPATITIS

steroids after 1 week of administration due to Alcoholic hepatitis is a clinical syndrome of
lack of treatment response.22 It predicts mor- jaundice and liver failure, often in the setting
tality rates within 6 months; a score of 0.45 of heavy alcohol use for decades.11,12 The inci-
or less indicates a good prognosis, and corti- dence is unknown, but the typical age of pre-
costeroid therapy should continue for 28 days sentation is between 40 and 50.11,12 The chief
(Figure 3).22 sign is a rapid onset of jaundice (< 3 months);
Our patient’s discriminant function score common signs and symptoms include fever,
is 50, her Glasgow score is 10, and her MELD ascites, proximal muscle loss, and an enlarged,
score is 28; thus, she begins treatment with oral tender liver.12 Encephalopathy may be seen in
prednisolone. Her Lille score at 1 week is 0.119, severe alcoholic hepatitis.12
indicating a good prognosis, and her corticoste- Our patient is 35 years old. She has jaun-
roids are continued for a total of 28 days. dice with rapid onset, as well as ascites and a
It should be highlighted that the most tender liver.
important treatment is abstinence from alco- The diagnosis of alcoholic hepatitis must
hol.11 Recent literature suggests that any ben- take into account the patient’s history, physi-
efit of prednisolone or pentoxifylline in terms cal examination, and laboratory findings. Un-
of mortality rates is questionable,19–20 and til proven otherwise, the diagnosis should be
there is evidence that giving both drugs simul- presumed in the following scenario: ascites
taneously may improve mortality rates,11,21 but and jaundice on examination (usually with a
the evidence remains conflicting at this time. duration < 3 months); a history of heavy al-
CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 86 • NUMBER 4 APRIL 2019 261

Downloaded from www.ccjm.org on April 26, 2022. For personal use only. All other uses require permission.
 NEW-ONSET ASCITES

cohol use; neutrophilic leukocytosis; an AST toms and laboratory testing results including
level that is elevated but below 300 U/L; an bilirubin improve. Her Lille score at 7 days in-
ALT level above the normal range but below dicates a good prognosis, prompting continu-
300 U/L; an AST-ALT ratio greater than 2; a ation of corticosteroid treatment for the full
total serum bilirubin level above 5 mg/dL; and 28 days.
an elevated INR.11,12 Liver biopsy is the gold She is referred to an outpatient alcohol re-
standard for diagnosis. Though not routinely habilitation program and has remained sober
done because of risks associated with the pro- as of the last outpatient note.
cedure, it may help confirm the diagnosis if it Alcoholic hepatitis is extremely difficult to
is in question. diagnose, and no single blood test or imaging
study confirms the diagnosis. The history, phys-
■ CASE CONCLUDED ical examination findings, and laboratory find-
We start our patient on oral prednisolone 40 ings are crucial. If the diagnosis is still in doubt,
mg daily for alcoholic hepatitis. Her symp- liver biopsy may help confirm the diagnosis. ■
■ REFERENCES 13. Maddrey WC, Boitnott JK, Bedine MS, Weber FL Jr, Mezey E, White
RI Jr. Corticosteroid therapy of alcoholic hepatitis. Gastroenterology
1. Ruyon BA; AASLD Practice Guidelines Committee. Management of 1978; 75(2):193–199. pmid:352788
adult patients with ascites due to cirrhosis: an update. Hepatology 14. Forrest EH, Evans CD, Stewart S, et al. Analysis of factors predictive
2009; 49(6):2087–2107. doi:10.1002/hep.22853 of mortality in alcoholic hepatitis and derivation and validation of
2. Hoefs JC, Canawati HN, Sapico FL, Hopkins RR, Weiner J, Mont- the Glasgow alcoholic hepatitis score. Gut 2005; 54(8):1174–1179.
gomerie JZ. Spontaneous bacterial peritonitis. Hepatology 1982; doi:10.1136/gut.2004.050781
2(4):399–407. pmid:7095741 15. Dunn W, Jamil LH, Brown LS, et al. MELD accurately predicts mortal-
3. Ginès P, Cárdenas A, Arroyo V, Rodés J. Management of cirrhosis ity in patients with alcoholic hepatitis. Hepatology 2005; 41(2):353–
and ascites. N Engl J Med 2004; 350(16):1646–1654. 358. doi:10.1002/hep.20503
doi:10.1056/NEJMra035021 16. Sheth M, Riggs M, Patel T. Utility of the Mayo end-stage liver dis-
4. Runyon BA, Montano AA, Akriviadis EA, Antillon MR, Irving MA, ease (MELD) score in assessing prognosis of patients with alcoholic
McHutchison JG. The serum-ascites albumin gradient is superior hepatitis. BMC Gastroenterol 2002; 2:2. pmid:11835693
to the exudate-transudate concept in the differential diagnosis of 17. Akriviadis E, Botla R, Briggs W, Han S, Reynolds T, Shakil O. Pent-
ascites. Ann Intern Med 1992; 117(3):215–220. pmid:1616215 oxifylline improves short-term survival in severe acute alcoholic
5. Hernaez R, Hamilton JP. Unexplained ascites. Clin Liver Dis hepatitis: a double-blind, placebo-controlled trial. Gastroenterology
2016; 7(3):53–56. https://aasldpubs.onlinelibrary.wiley.com/doi/ 2000; 119(6):1637–1648. pmid:11113085
epdf/10.1002/cld.537 18. Mathurin P, O’Grady J, Carithers RL, et al. Corticosteroids improve
6. Huang LL, Xia HH, Zhu SL. Ascitic fluid analysis in the differential short-term survival in patients with severe alcoholic hepatitis:
diagnosis of ascites: focus on cirrhotic ascites. J Clin Transl Hepatol meta-analysis of individual patient data. Gut 2011; 60(2):255–260.
2014; 2(1):58–64. doi:10.14218/JCTH.2013.00010 doi:10.1136/gut.2010.224097
7. Bartoloni A, Zammarchi L. Clinical aspects of uncomplicated and 19. Thursz MR, Richardson P, Allison M, et al; STOPAH Trial. Predniso-
severe malaria. Mediterr J Hematol Infect Dis 2012; 4(1):e2012026. lone or pentoxifylline for alcoholic hepatitis. N Engl J Med 2015;
doi:10.4084/MJHID.2012.026 372(17):1619–1628. doi:10.1056/NEJMoa1412278
8. Titó L, Rimola A, Ginès P, Llach J, Arroyo V, Rodés J. Recurrence of 20. Thursz M, Forrest E, Roderick P, et al. The clinical effectiveness and
spontaneous bacterial peritonitis in cirrhosis: frequency and predic- cost-effectiveness of steroids or pentoxifylline for alcoholic hepatitis
tive factors. Hepatology 1988; 8(1):27–31. pmid:3257456 (STOPAH): a 2 × 2 factorial randomised controlled trial. Health Tech-
9. Fernández J, Ruiz del Arbol L, Gómez C, et al. Norfloxacin vs ceftri- nol Assess 2015; 19(102):1–104. doi:10.3310/hta191020
axone in the prophylaxis of infections in patients with advanced cir-
21. Lee YS, Kim HJ, Kim JH, et al. Treatment of severe alcoholic hepatitis
rhosis and hemorrhage. Gastroenterology 2006; 131(4):1049–1056.
with corticosteroid, pentoxifylline, or dual therapy: a systematic
doi:10.1053/j.gastro.2006.07.010
review and meta-analysis. J Clin Gastroenterol 2017; 51(4):364–377.
10. Runyon B; The American Association for the Study of Liver Diseases
doi:10.1097/MCG.0000000000000674
(AASLD). Management of adult patients with ascites due to cir-
22. Louvet A, Naveau S, Abdelnour M, et al. The Lille model: a new
rhosis: update 2012. https://www.aasld.org/sites/default/files/guide-
tool for therapeutic strategy in patients with severe alcoholic
line_documents/141020_Guideline_Ascites_4UFb_2015.pdf. Accessed
hepatitis treated with steroids. Hepatology 2007; 45(6):1348–1354.
September 4, 2018.
doi:10.1002/hep.21607
11. Sidhu SS, Goyal O, Kishore H, Sidhu S. New paradigms in manage-
ment of alcoholic hepatitis: a review. Hepatol Int 2017; 11(3):255– ADDRESS: William C. Lippert, MD, MPH, Section of Hospital Medicine,
267. doi:10.1007/s12072-017-9790-5 Department of Internal Medicine, Wake Forest Baptist Medical Center, 1
12. Lucey MR, Mathurin P, Morgan TR. Alcoholic hepatitis. N Engl J Med Medical Center Boulevard, Winston-Salem, NC 27157;
2009; 360(26):2758–2769. doi:10.1056/NEJMra0805786 [email protected]

262 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 86 • NUMBER 4 APRIL 2019

Downloaded from www.ccjm.org on April 26, 2022. For personal use only. All other uses require permission.