Use of noninvasive markers of portal hypertension and timing of screening endoscopy for gastroesophageal varices in patients with chronic liver disease


  • Annalisa Berzigotti,

    1. Hepatic Hemodynamic Laboratory Liver Unit, Hospital Clinic, Institut d'Investigacions Biomediques August Pi i Sunyer (IDIBAPS), University of Barcelona, Barcelona, Spain
    2. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBERehd), University of Barcelona, Barcelona, Spain
    Search for more papers by this author
  • Jaime Bosch,

    Corresponding author
    1. Hepatic Hemodynamic Laboratory Liver Unit, Hospital Clinic, Institut d'Investigacions Biomediques August Pi i Sunyer (IDIBAPS), University of Barcelona, Barcelona, Spain
    2. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBERehd), University of Barcelona, Barcelona, Spain
    • Address reprint requests to: Jaime Bosch, M.D., Ph.D., Hepatic Hemodynamic Laboratory, Liver Unit, Hospital Clínic, c/Villarroel 170, 08036 Barcelona, Spain. E-mail:; fax: +34 93 227 93 48.

    Search for more papers by this author
  • Thomas D. Boyer

    1. Department of Medicine, University of Arizona College of Medicine, Tucson, AZ
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

  • CIBERehd is funded by Instituto de Salud Carlos III. J.B. is the recipient of a grant from ISCiii (PS09/01261).


clinically significant portal hypertension


esophageal varices


gastroesophageal varices


hepatic vein pressure gradient


spleen size/platelet count


portal hypertension


transient elastography.

Editors' Note

Clinical perspectives in hepatology aims to engage two experts with opinions supporting differing perspectives on the management of a case. Typically, the case represents an area of debate or evolving practice in clinical hepatology.

The case described by Drs. Berzigotti and Bosch provides an opportunity to discuss challenges in the timing of endoscopy in an era of increasing availability of noninvasive methods to diagnose portal hypertension (PH).

Drs. Berzigotti and Bosch: Case Presentation

A 63-year-old woman with posttransfusion chronic hepatitis C (genotype 1b; low viremia; interleukin 28B rs12979860 CC genotype) lasting over 25 years was referred to our outpatient unit to evaluate eligibility for antiviral treatment. Laboratory tests indicated minimally elevated alanine aminotransferase (57 U/L), mildly increased international normalized ratio (1.13), and mild thrombocytopenia (127,000/mm3). Bilirubin and albumin were normal (0.8 mg/dL and 40 g/L, respectively). Abdominal ultrasound pointed out nodular liver surface and mild splenomegaly (bipolar diameter: 13 cm), further suggesting the presence of cirrhosis.[1] Liver stiffness by transient elastography (TE) was 15.6 kPa. Because this measurement was performed in the afternoon in nonfasting conditions,[2] it was repeated the next morning after fasting overnight, resulting in 9.6 kPa. To better assess the risk of PH and esophageal varices (EV), we used the combination of platelet count, spleen size, and liver stiffness according to recent data from our group and others.[3, 4] PH risk score was −1.41, corresponding to a 19.6% probability of having clinically significant portal hypertension (CSPH; defined by hepatic vein pressure gradient [HVPG] ≥10 mmHg). Varices risk score was −2.80, indicating a 5.8% probability of having EV. Liver stiffness × spleen size/platelet count (LSPS) was 0.98, also suggesting the absence of EV.

Therefore, noninvasive methods suggested that this patient had cirrhosis, but a low risk of having CSPH and a minimal risk of having varices. To provide robust data in order to select the best antiviral therapy[5, 6] and comply with international recommendations,[7] liver biopsy, HVPG measurement, and upper digestive tract endoscopy were performed. Liver biopsy confirmed cirrhosis, HVPG was slightly elevated (7 mmHg), and no gastroesophageal varices (GOV) were found on endoscopy.

Drs. Berzigotti and Bosch's Perspective

According to accepted international recommendations,[7] all patients with newly diagnosed cirrhosis should undergo screening endoscopy for assessing GOV in order to begin primary prophylaxis, if required, and HVPG measurement should be obtained for prognostic aims whenever available. However, thanks to improvements in noninvasive methods to quantify liver fibrosis,[8] at present most patients are diagnosed in a very initial stage of cirrhosis, in which CSPH and varices are often absent.[3] In this new scenario, a large proportion of HVPG measurements and screening endoscopy are unnecessary. Therefore, efforts should be directed at limiting these procedures to those patients at higher risk of CSPH and varices, so as to reducing healthcare cost and lessen patients' discomfort. Recently, new, simple, noninvasive tests based on liver elastography, alone or in combination with LSPS size, or on spleen stiffness have been described (Table 1). LSPS showed very similar accuracy across independent studies. This combination, or the ones we described based on more robust statistical modeling,[3] can be of great help in assisting clinical decision making in this field. Cutoffs with 90% sensitivity and 90% specificity can be chosen, allowing for reliably ruling out and diagnosing CSPH or varices. It appears reasonable to spare HVPG measurement and endoscopy in patients with <20% probability of CSPH based on the combination of noninvasive tests. In our experience, in 173 patients, only 3 of 70 with varices (4%; all with small varices) would have been missed if endoscopy was delayed using these criteria.[3] According to our data, the timing of first screening endoscopy can be safely postponed until noninvasive tests indicate the presence of CSPH.

Table 1. Recently Described Noninvasive Methods to Identify PH and EV in Patients With Compensated Cirrhosis
 NPrediction of CSPH (HVPG ≥10 mmHg)Prediction of EV
  1. N refers to the sum of patients studied across published reports.

  2. Abbreviations: ARFI, acoustic radiation force impulse imaging; AUROC, area under the receiver operating characteristic curve; N/A, not available.

Liver stiffness by TE[9, 10]>4000.76-0.990.71-0.89
Liver stiffness by ARFI[11]340N/A0.74
Spleen stiffness by TE[10, 12]100 + 1350.97; N/A0.94; 0.78
Spleen stiffness by ARFI[11]340N/A0.93
LSPS[3, 4, 10]117 + 56 + 100 + 2800.91-0.920.88-0.95
PH risk score[3]117 + 560.94;0.93-
Varices risk score[3]117 + 56-0.91; 0.76

Dr. Boyer's Perspective

Drs. Berzigotti and Bosch have very nicely shown the value of different tests in determining which patients with cirrhosis are at risk for varices. These tests could be useful in selecting patients with cirrhosis likely to benefit from endoscopy to confirm and grade varices. There are two issues that need to be discussed before embracing elastography plus LSPS as the best approach to screening. The addition of a new test such as elastography could add costs that may exceed that of a screening endoscopy. To determine effectiveness, some clinical endpoint, such as bleeding, needs to be included to determine cost-effectiveness. The finding of no or small varices is of unclear benefit because we lack treatments that either prevent the appearance of varices or slow their growth. A better goal is identification of large varices for which we have therapeutic options of proven benefit.

Thus, if our goal is to find high-risk varices, then no current tests, other than endoscopy, will identify these patients reliably. A previous study found that 28% of patients with cirrhosis with a platelet count of <88,000 or splenomegaly, compared to 7% in those who lacked these features, had large varices. There was a significant cost savings when only those at greatest risk for large varices underwent endoscopy.[13] However, 7% of patients would have undiagnosed high-risk varices, which, in my opinion, is an unacceptably high number given the consequences of a variceal bleed and the availability of effective preventative therapies. We can reduce the false-negative rate to near zero using elastography and LSPS, but only ∼18% of this group will have large varices and an even smaller number will bleed.[3] Is this cost-effective relative to endoscoping all patients? Given the rising cost of healthcare, I believe we need to move away from the do-it-all approach and be more measured in our care of patients. But, to make intelligent choices, we need good cost-effectiveness data, which we currently lack.