Noninvasive assessment of hepatic decompensation

Abstract

Noninvasive tests (NITs) are used in all aspects of liver disease management. Their most prominent break-through since the millennium has been in advancing early detection of liver fibrosis, but their use is not limited to this. In contrast to the symptom-driven assessment of decompensation in patients with cirrhosis, NITs provide not only opportunities for earlier diagnoses but also accurate prognostication, targeted treatment decisions, and a means of monitoring disease. NITs can inform disease management and decision-making based on validated cutoffs and standardized interpretations as a valuable supplement to clinical acumen. The Baveno VI and VII consensus meetings resulted in tangible improvements to pathways of care for patients with compensated and decompensated advanced chronic liver disease, including the combination of platelet count and transient elastography to diagnose clinically significant portal hypertension. Furthermore, circulating NITs will play increasingly important roles in assessing the response to interventions against ascites, variceal bleeding, HE, acute kidney injury, and infections. However, due to NITs' wide availability, there is a risk of inaccurate use, leading to a waste of resources and flawed decisions. In this review, we describe the uses and pitfalls of NITs for hepatic decompensation, from risk stratification in primary care to treatment decisions in outpatient clinics, as well as for the in-hospital management of patients with acute-on-chronic liver failure. We summarize which NITs to use when, for what indications, and how to maximize the potential of NITs for improved patient management.

Graphical abstract

FIGURE 1

Bayes theorem exemplified by transient elastography for diagnosing advanced alcohol-related fibrosis in primary versus secondary care. Prevalences and the sensitivity and specificity of transient elastography are taken from Thiele et al. The prior probability denotes the risk of advanced fibrosis before having the test result. In a diagnostic setting, the prior probability equals the prevalence; in a prognostic setting, it equals the incidence. Posterior probabilities denote the risk assessment after the test result is obtained. In a diagnostic setting, this is similar to the positive and negative predictive values. In the example, the risk for advanced fibrosis in a patient from primary care with liver stiffness below 15 kPa is calculated as: $1\mathrm{–}\mathrm{NPV}=1\mathrm{–}\left(\frac{\mathit{specificity}\times (1-\mathit{prevalence})}{\mathit{specificity}\times \left(1-\mathit{prevalence}\right)+(1-\mathit{sensitivity})\times \mathit{prevalence}}\right)=1\mathrm{–}\left(\frac{0.95\times 0.94}{0.95\times 0.94+0.09\times 0.06}\right)=1\mathrm{–}0.994=0.6\%$ . The posterior probability is consequently highly dependent on both test accuracy and disease prevalence. From this example, it is also clear that if transient elastography is performed in primary care, a result of 15.2 kPa should be interpreted differently from the same result for the same patient in secondary care. In primary care, slightly more than half of patients with LSM ≥ 15 kPa have advanced fibrosis, compared to 91% of patients in secondary care. Knowledge of predisposing factors that influence prevalence will also affect posterior probability. Examples from Israelsen et al. Abbreviation: LSM, liver stiffness measurement (transient elastography).

FIGURE 2

(A) Natural history of chronic liver disease. (B) NITs to be used according to the clinical spectrum of chronic liver disease. NITs for liver disease begin with diagnosing liver fibrosis and hepatic inflammation, progress to identification of portal hypertension, and culminate with monitoring complications and their treatment, organ failure, and death. Abbreviations: ACLF, acute-on-chronic liver failure; cACLD, compensated advanced chronic liver disease; CSPH, clinically significant portal hypertension; NIT, noninvasive test.

FIGURE 3

(A) Risk of hepatic decompensation stratified by baseline fibrosis stage and etiology. (B) Risk of hepatic decompensation stratified by baseline transient elastography value and etiology. Adapted from Rasmussen et al. and Boursier et al