Skip to content

Hepatocellular carcinoma — staging and treatment allocation

TL;DR — HCC staging must describe both cancer and liver because the same tumor has different prognosis and treatment feasibility in compensated versus failing liver. BCLC 2022 combines tumor burden, liver function, performance status, and treatment options into a five-stage strategy, and is the dominant Western allocation framework (Reig 2022, PMID 34801630). It is a decision scaffold rather than an immutable algorithm: treatment-stage migration, technical feasibility, transplant policy, competing mortality, and patient priorities frequently justify movement across nominal stage-linked therapies (Vitale 2020, PMID 32064645). Alternative systems may discriminate prognosis better in selected regional cohorts, but no staging system is globally accepted across heterogeneous populations (Tellapuri 2018, PMID 30593649). Multidisciplinary review is therefore a substantive part of staging, not an administrative afterthought.

Why conventional TNM is insufficient

In many solid tumors, anatomic extent dominates stage. In HCC, survival and treatment depend simultaneously on:

  • tumor number, size, distribution, and biology;
  • macrovascular invasion and extrahepatic spread;
  • portal hypertension and liver reserve;
  • performance status and comorbidity;
  • transplant eligibility and organ availability;
  • technical feasibility and local expertise.

TNM remains useful for pathology and surgical series, but it does not by itself allocate therapy in cirrhosis.

BCLC 2022 at a glance

Stage Typical tumor/liver context Default treatment direction Major branch points
BCLC 0 Solitary very early tumor; preserved function; PS 0 Ablation or resection Location, portal hypertension, transplant strategy
BCLC A Solitary or limited multifocal early HCC; preserved function Resection, ablation, or transplantation Number/size, liver reserve, transplant eligibility
BCLC B Multinodular disease without vascular invasion/spread; PS 0 TACE for suitable disease; transplant/systemic options in selected subgroups Tumor burden, arterial anatomy, liver function, downstaging
BCLC C Vascular invasion or extrahepatic spread; preserved enough function; PS 1–2 Systemic therapy Bleeding risk, autoimmune/transplant context, liver reserve
BCLC D End-stage liver function or PS 3–4 not attributable to treatable tumor Symptom-directed/supportive care; transplant only in exceptional eligible context Reversibility, transplant candidacy, goals

This table compresses branches in the full strategy. BCLC 2022 introduced explicit concepts of individualized clinical decision-making, treatment-stage migration, and “untreatable progression” rather than equating one stage with one procedure (Reig 2022, PMID 34801630).

Tumor burden

Minimum tumor descriptors are:

  • number of viable lesions;
  • largest viable diameter;
  • unilobar or bilobar distribution;
  • segmental anatomy and proximity to vessels/bile ducts;
  • macrovascular invasion, including branch and main portal vein;
  • extrahepatic lymph-node or distant spread;
  • AFP and rate of change where relevant;
  • radiographic response to previous therapy.

Tumor burden is continuous. Dichotomies such as “within” or “beyond” Milan criteria are useful for allocation but discard biological information near either side of the boundary.

Liver function

Child-Pugh remains embedded in trials and practice, but contains subjective variables (ascites and encephalopathy), arbitrary cut points, and correlated laboratory components. Albumin-bilirubin (ALBI) uses two objective laboratory values and can subdivide apparently well-compensated populations.

Measure Components Strength Limitation
Child-Pugh Albumin, bilirubin, INR, ascites, encephalopathy Familiar; trial eligibility and clinical context Subjective components; ceiling/floor effects
MELD/MELD-Na Bilirubin, INR, creatinine ± sodium Short-term liver mortality and allocation Not an HCC tumor model
ALBI Albumin, bilirubin Objective and continuous Omits portal hypertension and clinical decompensation
Modified ALBI Subdivides ALBI grade 2 More granularity in preserved-function HCC Regional validation and cut-point dependence (Hiraoka 2019, PMID 31019902)

A multicentre Japanese analysis validated modified ALBI grading for finer assessment of hepatic reserve in HCC, especially within the broad ALBI grade 2 population (Hiraoka 2019, PMID 31019902). Laboratory scores should complement, not erase, ascites, encephalopathy, variceal bleeding, frailty, and transplant status.

Performance status

Performance status predicts prognosis and treatment tolerance, but attribution matters. A high score caused by cancer burden is different from disability caused by cirrhosis, arthritis, frailty, or social circumstance. Mechanical insertion of ECOG status can upstage patients whose limitation is not tumor-driven.

Document:

  • baseline function before recent decompensation;
  • whether limitation is tumor-, liver-, or comorbidity-driven;
  • reversibility after ascites, infection, anemia, or pain treatment;
  • nutritional and frailty assessment;
  • patient-defined functional priorities.

Portal hypertension and resection allocation

Clinically significant portal hypertension increases postoperative decompensation risk. Platelets, splenomegaly, varices, elastography, and direct pressure measurement provide overlapping evidence. A small resectable tumor in a patient with major portal hypertension may be better allocated to ablation or transplant than resection (Singal 2023, PMID 37199193).

Conversely, rigid exclusion by a single surrogate can deny resection to carefully selected patients. Allocation should estimate future liver remnant, location, surgical complexity, portal hypertension, and salvage-transplant options together.

Treatment-stage migration

Treatment-stage migration permits use of the next appropriate therapy when the nominal first option is unsuitable or has failed. Examples include:

  • ablation instead of resection for a small tumor in a high-risk surgical patient;
  • TACE or radiation for an early tumor inaccessible to safe ablation;
  • systemic therapy for diffuse BCLC B disease unlikely to benefit from embolization;
  • transplant after successful downstaging from beyond conventional criteria;
  • locoregional therapy for selected limited vascular invasion in expert centres.

Vitale and colleagues distinguish a rigid stage hierarchy from a therapeutic hierarchy that ranks potentially effective treatments for an individual rather than linking each stage to only one treatment (Vitale 2020, PMID 32064645).

Migration is not license for arbitrary escalation. It should state:

  1. why the stage-default option is infeasible or inferior;
  2. what evidence supports the substitute;
  3. what liver-function boundary applies;
  4. whether intent is curative, bridging, downstaging, or palliative.

BCLC B is deliberately heterogeneous

Intermediate-stage HCC ranges from a few transplantable lesions to diffuse bilobar disease. The BCLC 2022 strategy separates patients by whether they are candidates for transplant, have well-defined nodules with preserved portal flow suitable for selective TACE, or have infiltrative/extensive disease better served by systemic therapy (Reig 2022, PMID 34801630).

BCLC B feature Allocation implication
Within transplant/downstaging policy Evaluate transplant pathway
Discrete arterialized nodules, preserved portal flow Selective TACE may provide durable control
Diffuse/infiltrative bilobar burden Low probability of TACE benefit; consider systemic therapy
Worsening liver function after repeated TACE Stop ineffective repetition; migrate therapy
High AFP or rapid progression Signals adverse biology and need to reassess intent

Alternative staging systems

System Main emphasis Use/caution
TNM Anatomic/pathologic extent Surgical pathology; omits liver function
Okuda Tumor proportion and liver function Historical; coarse for modern early disease
CLIP Child-Pugh, morphology, AFP, portal-vein thrombosis Prognostic; less directly therapeutic
HKLC More granular tumor/liver groups and aggressive treatment allocation Developed in HBV-predominant Asian cohort
ITA.LI.CA Prognostic integration Regional derivation
MESIAH Continuous prognostic score Prognosis rather than direct treatment map
BCLC Prognosis plus treatment allocation Widely used; can be over-rigidly interpreted

A review concluded that no single globally accepted staging system permits straightforward comparison across heterogeneous populations (Tellapuri 2018, PMID 30593649). In a 668-patient Chinese cohort, HKLC showed higher time-specific AUCs than BCLC (for example 1-year AUC 0.740 versus 0.622), but this single-region comparison does not establish universal superiority (Yan 2015, PMID 26067223).

Transplant selection overlays stage

Milan criteria—one tumor ≤5 cm or up to three each ≤3 cm without macrovascular invasion or extrahepatic spread—remain a central transplant benchmark (Mazzaferro 1996, PMID 8594428). Modern allocation may also use AFP, response to downstaging, waiting time, and regional policy.

Metroticket 2.0 combines AFP with tumor number and size to predict HCC-specific death after transplant and outperformed purely morphologic criteria in its development/validation cohorts (Mazzaferro 2018, PMID 28989060). This illustrates the difference between staging, prognosis, and scarce-organ allocation: they overlap but are not identical.

Multidisciplinary allocation

A useful HCC board includes hepatology, transplant and hepatobiliary surgery, interventional radiology, diagnostic radiology, medical oncology, radiation oncology, pathology, and supportive care as needed.

The board should output more than a stage label:

Output Required content
Diagnosis confidence Imaging category, pathology status, unresolved differential
Tumor extent Number, size, distribution, vascular invasion, spread
Liver state Decompensation history, portal hypertension, Child-Pugh/MELD/ALBI
Candidate therapies Ranked options and contraindications
Intent Curative, bridging, downstaging, disease control, symptom control
Reassessment point Imaging interval and stopping/migration rule
Trial option Stage- and liver-function-compatible studies

Re-staging over time

HCC stage is dynamic. Reassess after:

  • locoregional therapy;
  • systemic response or progression;
  • new vascular invasion or metastasis;
  • decompensation or improvement in liver function;
  • successful downstaging;
  • transplant-list change;
  • emergence of a new primary tumor in the cirrhotic field.

Record both baseline and current stage. Post-treatment prognosis depends on trajectory, not only the latest cross-sectional category.

Common allocation errors

  • Using tumor anatomy without liver function.
  • Using Child-Pugh class without decompensation history.
  • Treating BCLC as a mandatory one-stage/one-treatment rule.
  • Repeating TACE after loss of selectivity or liver reserve.
  • Calling disability tumor-related without attribution.
  • Confusing transplant eligibility with organ allocation priority.
  • Ignoring patient time, travel, and treatment-burden preferences.
  • Failing to define a stopping or reassessment rule.

Validation, migration, and real allocation

The original BCLC proposal linked stage to prognosis and treatment using tumor extent, liver function, and performance status (Llovet 1999, PMID 10518312). External validation supports prognostic separation, but early-stage simplifications and regional systems can recalibrate survival in cohorts selected for surgery or HBV-predominant disease (Santambrogio 2013, PMID 23726257; Yang 2023, PMID 36167767). The practical question is therefore not whether BCLC “works,” but when its default allocation should yield to a better individualized option.

Allocation problem Quantified evidence Operational implication
Liver reserve metric Systematic review generally found ALBI at least comparable to Child-Pugh for outcome discrimination Use an objective continuous measure alongside, not in isolation from, decompensation and portal hypertension (Peng 2020, PMID 32240595)
Early-stage migration Australian multicentre data found high treatment-stage migration and adverse outcomes among patients not receiving stage-concordant therapy Record why migration occurred: anatomy, reserve, access, preference, or clinician strategy (Loo 2022, PMID 36091321)
Multidisciplinary divergence from BCLC In 321 patients, tumor-board recommendations frequently individualized beyond the algorithm Divergence can be rational, but should be auditable rather than implicit (Matsumoto 2021, PMID 33825060)
Adherence to tumor-board plan Long-term survival was associated with implementation of multidisciplinary recommendations A board without downstream access is not an intervention (Oehring 2025, PMID 40981711)
Portal hypertension and resection Meta-analyses associate clinically significant portal hypertension with higher perioperative and long-term risk Measurement method changes apparent risk; selected patients may still benefit (Choi 2014, PMID 24867654; Aliseda 2024, PMID 38126757)
Small solitary HCC Comparative data across resection, ablation, and SBRT show different local-control and toxicity profiles “BCLC A” does not specify the best local therapy without size, location, reserve, and salvage plan (Maher 2024, PMID 38400681)
Intermediate-stage TACE Tumor burden, AFP, ALBI, and treatment response stratify survival BCLC B is a family of risks, not a single prognosis (Ogasawara 2015, PMID 25919025; Xia 2025, PMID 39758510)

The migration decision

Treatment-stage migration can mean three different things: inability to deliver the nominal therapy, deliberate selection of a superior alternative, or transition after treatment failure. These have different prognostic meanings and should not be pooled (Wehling 2021, PMID 33537908). After TACE, radiologic stage migration correlates with survival, but it is partly a marker of tumor biology and liver injury rather than a validated surrogate endpoint for every treatment comparison (Kassab 2023, PMID 36972387).

For intermediate disease, repeated embolization should be conditional on response and preserved liver function. Scores such as ART, ABCR, and TACE-specific tumor-burden models have inconsistent external performance; EZ-ALBI offers a simple liver-reserve measure but does not capture arterial anatomy, infiltrative phenotype, or extrahepatic risk (Ananchuensook 2022, PMID 35701739; Elshaarawy 2019, PMID 31372364). A therapeutic hierarchy—ranking feasible options by expected survival benefit—may be more clinically useful than treating the stage label as a prohibition (Vitale 2020, PMID 32064645).

Controversies

  • Algorithm adherence versus personalization. Population-level adherence is associated with better outcomes, but rigid adherence can deny resection, transplant, radiation, or systemic therapy to selected patients who fall outside default boxes (Iavarone 2025, PMID 40658790; Matsumoto 2021, PMID 33825060).
  • Child-Pugh versus ALBI. ALBI removes subjective ascites and encephalopathy grading; Child-Pugh retains clinically meaningful decompensation. Neither captures sarcopenia, portal-pressure physiology, or competing comorbidity (Peng 2020, PMID 32240595).
  • Portal hypertension as contraindication versus risk factor. Higher risk is consistent, but modern minimally invasive surgery and careful selection challenge a categorical ban (Choi 2014, PMID 24867654; Aliseda 2024, PMID 38126757).
  • Stage migration as quality failure versus good care. Migration caused by access failure is harmful; migration chosen by an expert board may improve fit. Databases must distinguish the reason (Loo 2022, PMID 36091321; Oehring 2025, PMID 40981711).

Open questions

  • Does therapeutic hierarchy improve survival over guideline stage hierarchy without increasing selection bias (Vitale 2020, PMID 32064645)?
  • Which BCLC B features should prospectively select systemic therapy over first TACE (Reig 2022, PMID 34801630)?
  • Can ALBI, portal-hypertension measures, and frailty replace subjective liver-function classification in trials (Hiraoka 2019, PMID 31019902)?
  • How should response to downstaging be incorporated into a unified stage rather than a separate transplant policy?
  • Which staging system transports best across viral, alcohol, and MASLD populations (Tellapuri 2018, PMID 30593649)?

References

  1. Reig M, et al. BCLC strategy for prognosis prediction and treatment recommendation: the 2022 update. J Hepatol. 2022;76:681-693. PMID 34801630
  2. Vitale A, et al. Treatment of hepatocellular carcinoma in the precision medicine era: from treatment stage migration to therapeutic hierarchy. Hepatology. 2020;72:2206-2218. PMID 32064645
  3. Tellapuri S, et al. Staging systems of hepatocellular carcinoma: a review. Indian J Gastroenterol. 2018;37:481-491. PMID 30593649
  4. Hiraoka A, et al. Validation of modified ALBI grade for more detailed assessment of hepatic function in hepatocellular carcinoma patients: a multicenter analysis. Liver Cancer. 2019;8:121-129. PMID 31019902
  5. Yan X, et al. Validation of models in patients with hepatocellular carcinoma: comparison of HKLC with BCLC in a Chinese cohort. Eur J Gastroenterol Hepatol. 2015;27:1180-1186. PMID 26067223
  6. Singal AG, et al. AASLD Practice Guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatology. 2023;78:1922-1965. PMID 37199193
  7. Mazzaferro V, et al. Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med. 1996;334:693-699. PMID 8594428
  8. Mazzaferro V, et al. Metroticket 2.0 model for analysis of competing risks of death after liver transplantation for hepatocellular carcinoma. Gastroenterology. 2018;154:128-139. PMID 28989060
  9. Llovet JM, et al. Prognosis of hepatocellular carcinoma: the BCLC staging classification. Semin Liver Dis. 1999;19:329-338. PMID 10518312
  10. Santambrogio R, et al. External validation of a simplified BCLC staging system for early hepatocellular carcinoma. Eur J Surg Oncol. 2013;39:850-857. PMID 23726257
  11. Peng Y, et al. ALBI versus Child-Pugh in predicting outcome of patients with HCC: a systematic review. Expert Rev Gastroenterol Hepatol. 2020;14:383-400. PMID 32240595
  12. Loo KF, et al. High rates of treatment stage migration for early hepatocellular carcinoma and association with adverse outcomes. JGH Open. 2022;6:599-606. PMID 36091321
  13. Wehling C, et al. Treatment stage migration and treatment sequences in patients with HCC: drawbacks and opportunities. J Cancer Res Clin Oncol. 2021;147:2471-2481. PMID 33537908
  14. Maher AM, et al. Outcomes after resection, ablation, or SBRT for solitary treatment-naïve HCC ≤3 cm. Cancer Med. 2024;13:e6978. PMID 38400681
  15. Iavarone M, et al. The impact of BCLC recommendations on survival for patients with hepatocellular carcinoma. Hepatol Commun. 2025;9. PMID 40658790
  16. Kassab I, et al. Stage migration as a surrogate of survival in hepatocellular carcinoma treated with TACE. Hepatol Commun. 2023;7. PMID 36972387
  17. Oehring R, et al. Adherence to multidisciplinary tumor-board decisions determines long-term survival in HCC. Ann Surg. 2025. PMID 40981711
  18. Matsumoto MM, et al. Comparing real-world multidisciplinary tumor-board recommendations with the BCLC algorithm. Cardiovasc Intervent Radiol. 2021;44:1070-1080. PMID 33825060
  19. Aliseda D, et al. Impact of portal-hypertension assessment method on outcomes of HCC resection: a meta-analysis. Ann Surg. 2024;280:46-55. PMID 38126757
  20. Choi SB, et al. Influence of clinically significant portal hypertension on outcomes following hepatectomy for HCC: a meta-analysis. J Hepatobiliary Pancreat Sci. 2014;21:639-647. PMID 24867654
  21. Ogasawara S, et al. A prognostic score for intermediate-stage HCC treated with TACE. PLoS One. 2015;10:e0125244. PMID 25919025
  22. Ananchuensook P, et al. Validation and prognostic value of EZ-ALBI in intermediate-stage HCC treated with TACE. BMC Gastroenterol. 2022;22:295. PMID 35701739
  23. Elshaarawy O, et al. Intermediate-stage hepatocellular carcinoma: a summary review. J Hepatocell Carcinoma. 2019;6:105-117. PMID 31372364
  24. Xia D, et al. Tumor burden with AFP improves survival prediction for TACE-treated HCC. JHEP Rep. 2025;7:101216. PMID 39758510
  25. Yang SC, et al. Prospective validation of Eastern Staging after surgical resection for HCC. HPB (Oxford). 2023;25:81-90. PMID 36167767