← Library

Compensated Cirrhosis — Long-Term Management

GESA 2026 (Australian pathway) · Baveno VIII 2026 (international comparison) · Cancer Council Australia HCC surveillance 2023
Compiled Aug 2026
Verify doses locally
Prevent the first decompensation
Decision support only. “Compensated” means no current or previous ascites, variceal bleeding or overt hepatic encephalopathy; new decompensation moves the patient to the acute pathway. Confirm aetiology-specific care and surveillance with gastroenterology/liver.
1 Decide what disease is present, then whether CSPH is present

The useful fork is clinically significant portal hypertension (CSPH). It predicts first decompensation and changes treatment. A FibroScan number is only useful if the method and population fit the rule.

Compensated cirrhosis confirmed or strongly suspected → establish cause, baseline severity and portal-hypertension status
Direct evidence
Endoscopic varices or portosystemic collaterals on imaging establish CSPH, regardless of transient-elastography thresholds. HVPG ≥10 mmHg is definitive where measured.
Valid VCTE setting
Use the Australian GESA “rule of five” for alcohol-related disease, viral hepatitis or non-obese MASLD (BMI <30). Measure fasting and interpret with platelets. Baveno VIII offers a parallel international NIT framework below.
Invalid / uncertain setting
BMI ≥30 with MASLD: use ANTICIPATE-NASH or specialist assessment. Acute hepatitis, cholestasis, congestion and a recent meal can falsely raise stiffness. ARFI/2D-SWE cut-offs are not interchangeable with VCTE. GESA 2026
CSPH confirmed
Consider NSBB to prevent first decompensation; carvedilol is preferred if tolerated.
CSPH not demonstrated
Treat the cause and monitor. No portal-pressure drug merely because the label says cirrhosis.
Inconclusive / cannot use NSBB
Assess for high-risk varices using VCTE + platelets, spleen stiffness where available, or endoscopy.
2 VCTE rule of five and the action attached to it
Fasting VCTE LSMGESA 2026 interpretationPractical next step
<5 kPaNormal.If other data suggest chronic liver disease, reconcile the discordance rather than dismissing it.
5 to <10Cirrhosis unlikely.Treat the underlying liver disease; review the basis for a prior cirrhosis label.
10 to <15Possible cirrhosis.Corroborate with platelets, imaging and another NIT; biopsy only if uncertainty will change management.
15 to <20Cirrhosis confirmed.CSPH is not automatically present. Continue the portal-hypertension assessment.
20 to <25Cirrhosis + possible CSPH. CSPH confirmed if platelets <150 ×10⁹/L.If platelets are ≥150, treat as an indeterminate CSPH result rather than quietly relabelling it positive.
≥25 kPaCirrhosis + CSPH confirmed.Discuss carvedilol if no contraindication; routine screening endoscopy is unnecessary once a target NSBB dose is tolerated. GESA 2026
Two rules answer different questions. LSM <20 kPa + platelets ≥150 excludes high-risk varices with NPV >95%; it does not prove that CSPH is absent. If using this endoscopy-sparing rule, repeat VCTE and platelets annually. GESA 2026

Baveno VIII: cACLD classification

  • ContextAn international comparison, not a replacement for the Australian GESA pathway. When disease is defined by NITs, Baveno VIII prefers compensated advanced chronic liver disease (cACLD); cACLD and histological/clinical compensated cirrhosis are not interchangeable.
  • LSM<10 kPa rules out cACLD if there are no other clinical/imaging signs; 10–15 is suggestive; >15 is highly suggestive. Use fasting VCTE and the most recent reliable value.
  • Follow-upRepeat NIT assessment for CSPH every 12 months. An indeterminate result may be observed and reassessed rather than automatically treated as positive. Baveno VIII 2026

Baveno VIII: parallel CSPH rules

  • Rule outLSM ≤15 kPa + platelets ≥150 ×10⁹/L rules out CSPH (NPV >90%). This is distinct from the GESA endoscopy-sparing rule for high-risk varices.
  • EstablishCSPH probability ≥75% by ANTICIPATE/ANTICIPATE-NASH or NICER; or LSM ≥25 kPa in viral/alcohol-related cACLD or non-obese MASLD; or spleen stiffness at 100 Hz >55 kPa.
  • LimitsUse ANTICIPATE-NASH for MASLD with BMI ≥30. Method-specific spleen-stiffness thresholds should not be transferred between devices. Baveno VIII 2026
3 Baseline work-up and disease modification

Confirm stage and reserve

  • HistoryPrior ascites, GI bleed or HE; alcohol quantity; metabolic risk; medicines/supplements; family history; pruritus, thrombosis and cancer symptoms.
  • BaselineFBC/platelets, EUC/Cr/Na, LFT/albumin, INR; calculate Child–Pugh and MELD-Na. US with portal/hepatic vein assessment and focal-lesion surveillance.
  • PhenotypeNutrition/frailty, sarcopenia, BP/HR, splenomegaly, oedema; screen symptoms of covert HE, HPS and portopulmonary hypertension when clinically suggested.

Do not leave the cause as “cryptogenic”

  • CommonHBsAg/anti-HBc and HCV testing; alcohol assessment; metabolic/MASLD review; ferritin + transferrin saturation.
  • DirectedANA/ASMA/IgG; AMA/IgM; MRCP where cholestatic; caeruloplasmin/urine copper in younger or atypical disease; alpha-1 antitrypsin phenotype when indicated.
  • OverlapMore than one driver is common. Viral cure does not neutralise ongoing alcohol or metabolic injury.

Treat the driver

  • ViralSuppress HBV and cure HCV using current Australian guidance.
  • AlcoholAbstinence support, pharmacotherapy when appropriate, addiction care and nutrition. “Cut down” is a weak plan once cirrhosis is established.
  • MetabolicTreat diabetes, obesity, dyslipidaemia and cardiovascular risk without sacrificing muscle. No drug should be presented as a cirrhosis cure.
  • Immune / otherDisease-specific therapy for AIH, PBC, haemochromatosis and other causes with specialist input.
Cause control can regress fibrosis, but it does not erase cirrhosis surveillance. Continue HCC surveillance after HCV cure, HBV suppression or sustained alcohol abstinence unless a liver specialist has explicitly reclassified risk. “Recompensation” is reserved for someone who previously decompensated; it is not a synonym for uncomplicated compensated disease. Baveno VIII informational definition: cause removed/controlled, ascites resolved on imaging off diuretics, no clinical HE off lactulose/rifaximin/LOLA, no recurrent variceal bleeding, all sustained for >6 months, with liver function in the Child–Pugh A5/A6 range. Continue six-monthly hepatology follow-up and HCC surveillance. Stop NSBB only if CSPH resolution is demonstrated; Baveno VIII advises stopping antibiotic SBP prophylaxis after genuine recompensation. Baveno VIII 2026
4 Prevent the first decompensation

If CSPH is confidently established

  • CarvedilolStart 6.25 mg daily; titrate to 12.5 mg/day as 6.25 mg BD or 12.5 mg daily if tolerated. In hypertension, GESA permits up to 25 mg/day. GESA 2026
  • TargetsCarvedilol is titrated to dose and tolerance, not a rigid heart-rate target. Maintain SBP >90 / MAP ≥65; review HR, symptoms and renal function after initiation/titration.
  • AlternativePropranolol if carvedilol unsuitable: GESA target at least 40 mg BD, titrated toward HR ~60 while preserving perfusion.
  • Avoid / holdAKI, hypotension, clinically important bradycardia/heart block; assess asthma and other contraindications. Revisit after the intercurrent problem resolves.

Endoscopy logic

  • On target NSBBNo routine screening endoscopy is required solely to find varices.
  • GESA pathwayIf no NSBB and LSM <20 + platelets ≥150, defer endoscopy and repeat both annually. If either threshold fails, perform endoscopy or specialist spleen-stiffness assessment. GESA 2026
  • Baveno VIIIIf NSBB-ineligible, perform EGD unless LSM <20 + platelets ≥150 or spleen stiffness <40 kPa. If index EGD finds no varices, repeat at 2 years with an active aetiological driver or 3 years after removal/suppression. International comparison
  • High-risk varicesNSBB or EVL. If EVL chosen, repeat every 1–3 months until eradication, then surveillance.
  • New decompensationReassess with endoscopy regardless of prior NSBB strategy. GESA 2026
Evidence boundary: PREDESCI supports NSBBs in confidently diagnosed CSPH, but participants had invasive HVPG confirmation; 67 received propranolol and only 33 carvedilol. GESA 2026 reasonably extrapolates to validated non-invasive CSPH, but the trial does not justify treating every borderline FibroScan. Its NNT of 9 is useful only if the phenotype is right.
5 Surveillance and review calendar
WhenWhat to coverAction threshold
Every visitAsk about abdominal distension, oedema, melaena/haematemesis, confusion/sleep reversal, jaundice, infection, falls, alcohol, adherence and adverse effects. Weight, BP/HR, volume status and frailty signal.Any ascites, variceal bleed or overt HE = decompensated pathway. New jaundice needs urgent assessment for decompensation, obstruction, drug injury and infection.
Typically q3–6 moFBC/platelets, Na/K/Cr, LFT/albumin and INR; recalculate Child–Pugh/MELD-Na when changing. Interval depends on stability, aetiology and treatment.Falling albumin/platelets, rising INR/bilirubin/Cr or accelerating LSM should prompt liver review rather than waiting for the next routine appointment.
Every 6 moLiver ultrasound; add AFP according to the Australian/local surveillance pathway. Surveillance continues while the patient remains a candidate for HCC-directed treatment. Cancer Council 2023New/enlarging solid lesion ≥10 mm, rising AFP or inadequate repeated US → specialist-directed multiphase CT or contrast MRI; CT/MRI are not routine first-line screening.
AnnuallyNutrition/frailty and functional capacity; vaccination status; bone risk; cardiometabolic risk; alcohol relapse risk. Repeat VCTE + platelets annually if using them to spare endoscopy.Escalate dietitian/exercise support for sarcopenia or frailty; DEXA interval follows baseline and risk.
After treatment changeCheck BP/HR, symptoms, Na/K/Cr after NSBB or other haemodynamically active therapy; timing is usually within 1–2 weeks and after titration.Symptomatic hypotension, SBP ≤90, MAP <65, AKI or important bradycardia → hold/reduce and reassess.
6 Whole-patient care and avoidable harm

Food, muscle and bone

  • Protein1.2–1.5 g/kg ideal body weight/day; spread across meals. Never restrict protein to prevent HE.
  • TimingAvoid long fasts; add a late carbohydrate/protein snack. Use dietitian support early.
  • ExerciseProgressive aerobic + resistance work, adapted to frailty and portal-hypertension risk.
  • BoneBaseline DEXA and vitamin D/calcium assessment; repeat by result and risk, especially cholestatic disease, low BMI or steroid exposure.

Vaccines and medicines

  • VaccinesInfluenza, COVID-19, pneumococcal and zoster when eligible; vaccinate against HAV/HBV if non-immune, following the Australian Immunisation Handbook.
  • AvoidNSAIDs; unnecessary PPIs; unreviewed supplements; routine benzodiazepines and opioids where safer options exist.
  • StatinsCompensated cirrhosis is not a reason to withhold a statin with a cardiovascular indication. Monitor appropriately; do not prescribe one solely as an unproven portal-pressure treatment.
  • AnticoagulationAbnormal INR is not “auto-anticoagulation”. Indications and bleeding risk require individual assessment.

Refer or escalate

  • Liver serviceAll established cirrhosis needs a coherent specialist plan. Refer sooner for uncertain cause/stage, CSPH, rising MELD, frailty, PVT, focal lesion or pregnancy.
  • TransplantFirst decompensation, HCC, refractory portal-hypertension complication or worsening synthetic/renal function should trigger assessment where appropriate. Do not wait for terminal physiology.
  • SurgeryElective procedures require liver-aware risk assessment, preferably VOCAL-Penn plus multidisciplinary review. “Child A” does not make major surgery routine.
  • Patient planGive written red flags: haematemesis/melaena, new confusion, fever, rapidly increasing girth, oliguria or jaundice.
Primary source: GESA. Diagnosis and management of portal hypertension in cirrhosis (2026); Baveno VIII. Advancing Consensus in Portal Hypertension (2026); Cancer Council Australia NHMRC-approved HCC surveillance guideline (2023); BSG/BASL outpatient compensated cirrhosis guidance (2023); AASLD HCC practice guidance (2023).
Key trials / evidence: Villanueva et al., PREDESCI (Lancet 2019), NSBBs for HVPG-confirmed CSPH; Singal et al. (J Hepatol 2022), HCC surveillance meta-analysis; AASLD malnutrition/frailty guidance (Hepatology 2021) and GESA 2026 support protein 1.2–1.5 g/kg/day and minimising fasting.
Caveats: GESA remains the Australian practice anchor; Baveno VIII is displayed as a contemporaneous international comparison and has not been substituted for local pathways. Both frameworks are method-, aetiology- and BMI-dependent. PREDESCI was small and used HVPG, so benefit should not be overextended to uncertain NIT-defined CSPH. HCC surveillance mortality evidence in Western cirrhosis is largely observational, but Australian and international guidelines consistently recommend six-monthly surveillance for treatment-eligible cirrhosis. Aetiology-specific therapy changes quickly; use the dedicated current guideline.