The adult HCC roadmap as a working console. Pick a stage and the pathway answers with the standard, the reasonable alternative, what to avoid, and the grade behind each — every claim tied to the paper it came from. The liver-directed evidence matrix covers all four tumours.
Three questions in order. Every yes settles the stage; every no moves to the next question. Performance status and spread are therefore decided before tumour burden is ever asked. Choose any answer or any stage to see the full recommendation.
Five parameters, 1–3 points each. Class A 5–6 · Class B 7–9 · Class C 10–15.
In primary biliary cholangitis and primary sclerosing cholangitis the bilirubin bands are <4, 4–10 and >10 mg/dL. The roadmap also records ALBI and MELD 3.0; Child-Pugh alone does not drive allocation.
| Grade | Definition |
|---|---|
| 0 | Fully active, able to carry on all pre-disease performance without restriction. |
| 1 | Restricted in physically strenuous activity but ambulatory and able to carry out work of a light or sedentary nature — light house work, office work. |
| 2 | Ambulatory and capable of all self-care but unable to carry out any work activities; up and about more than 50% of waking hours. |
| 3 | Capable of only limited self-care; confined to bed or chair more than 50% of waking hours. |
| 4 | Completely disabled; cannot carry on any self-care; totally confined to bed or chair. |
| 5 | Dead. |
Oken et al., Am J Clin Oncol 1982. In this pathway ECOG 3–4 settles stage D, and cancer-related ECOG 1–2 is part of stage C — record whether the limitation is from the tumour or the liver, because that distinction decides whether transplant is still on the table.
CT/MRI LI-RADS v2018 diagnostic algorithm and the v2024 treatment-response algorithms. Applies only in the at-risk population — cirrhosis of any cause, chronic hepatitis B, or current or prior HCC. OPTN class 5 criteria for listing are a separate system.
| Nonrim APHE | Unequivocal enhancement in the arterial phase, in whole or in part, that is not rim-shaped. |
| Nonperipheral washout | Visually assessed reduction in enhancement relative to liver, in whole or in part, from an earlier to a later phase — not peripheral. |
| Enhancing capsule | Smooth, uniform, sharp border around most or all of the observation, unequivocally thicker or more conspicuous than around background nodules. |
| Threshold growth | ≥50% increase in size in ≤6 months. |
Size is the fifth criterion and sets the row of the table; the three features other than APHE are what “additional major features” counts.
| LR-1 | Definitely benign. |
| LR-2 | Probably benign. |
| LR-3 | Intermediate probability of malignancy. |
| LR-4 | Probably HCC. |
| LR-5 | Definitely HCC. |
| LR-M | Probably or definitely malignant, not HCC specific — targetoid appearance or other features favouring cholangiocarcinoma or a combined tumour. |
| LR-TIV | Tumour in vein: enhancing soft tissue within a vessel, whatever the parenchymal observation. |
| LR-NC | Not categorisable because of image degradation or omission. |
| Additional major features | Any size |
|---|---|
| 0 | LR-3 |
| ≥1 | LR-4 |
| Additional major features | <10 mm | 10–19 mm | ≥20 mm |
|---|---|---|---|
| 0 | LR-3 | LR-3 | LR-4 |
| 1 | LR-4 | LR-4 if capsule alone LR-5 if washout or threshold growth | LR-5 |
| ≥2 | LR-4 | LR-5 | LR-5 |
An observation under 10 mm cannot be LR-5. LR-5 requires ≥10 mm with nonrim APHE plus nonperipheral washout or threshold growth, or ≥20 mm with nonrim APHE plus any one additional feature.
| Category | Nonradiation — ablation, TACE | Radiation — SBRT, Y-90 |
|---|---|---|
| LR-TR Nonviable | No lesional enhancement, or treatment-specific expected enhancement. | No lesional enhancement. |
| LR-TR Equivocal | Enhancement atypical for treatment effect and not meeting viable criteria. | Does not exist in the radiation algorithm. |
| LR-TR Nonprogressing | — | Unique to the radiation algorithm. Persistently enhancing, stable or decreasing — treatment effect rather than residual tumour. Observe with close follow-up; do not retreat on enhancement alone. |
| LR-TR Viable | Nodular, mass-like or thick irregular enhancement in or along the treated lesion. | New or enlarging mass-like enhancement. |
Applying the nonradiation algorithm after radioembolisation or SBRT is a recurring and consequential error in this pathway — it turns expected delayed necrosis into a false call of residual tumour.
Provenance. This table is standard reference material, not taken from the roadmap, which cites the ACR documents without reproducing them. Verify against the ACR CT/MRI LI-RADS v2018 core and the v2024 treatment-response algorithms before it is used for reporting.
Stage migration is permitted when anatomy, biology, liver function, transplant strategy, comorbidity, access or patient goals make the stage default unsuitable — the reason must be recorded. It is not a licence to substitute a lower-evidence modality for a proven curative option.
Intent is assigned before a device is chosen. Curative intent means an attempt to eradicate all known cancer; it does not establish a cure rate. Local control, complete radiographic response, complete pathologic necrosis in a treated lesion, and overall survival are four different outcomes.
Do not use a universal ablation → SBRT → Y-90 → TACE → HAI → histotripsy sequence. An upfront alternative chosen because of anatomy or organ reserve is not a post-progression second line. Bridge and downstaging preserve a later transplant or resection opportunity; conversion treatment is not itself the curative operation.
Compare resection, local ablation and transplant where appropriate at multidisciplinary review, and record the intended outcome, why alternatives fit or do not fit, the systemic-treatment plan, the reassessment date and the next action. After progression or toxicity, restage and reconsider complete treatment rather than automatically repeating the same procedure.
The tiers below are a clinical synthesis of treatment intent. They are not the roadmap's A–D evidence grading system, and the two should not be read as equivalent.
Loading the matrix…
Source: LRT_Evidence_Matrix, Batch 09 HAI update (15 October 2026), filtered to the hepatocellular carcinoma column. The full matrix also covers iCCA, colorectal and neuroendocrine liver metastases; perihilar cholangiocarcinoma is not in it.
Diagnosis and staging, the curative pathways and the CSPH gate, transplant and allocation, the locoregional selector, systemic therapy, and the response loop.
Retained because they are defensible practice here, but each rests on anatomy, physiology or consensus rather than comparative outcome data. Each requires prospective outcome tracking, and none should be quoted to a patient, a payer or a referring clinician as evidence based.
| Statement | What is actually known | Variables that should govern selection | Tracking commitment |
|---|
Sections 9, 10 and 12 of the roadmap, plus the ratification list. These are stakeholder-derived operating requirements — evidence grade D by construction, local policy choices with no comparative outcome evidence. They are labelled Standard because the programme chose them, not because trials support them, and should be audited for adherence rather than effectiveness until the registry can support an effectiveness question.
Every source behind the pathway, with the local depot naming and a link straight into the TMC library resolver. This is a decision-focused evidence synthesis, not a systematic review; single-arm response data are never used here to infer superiority.