Overview
Hepatocellular carcinoma (HCC) is the sixth most common cancer globally and the second leading cause of cancer-related death, accounting for approximately 830,000 deaths annually. Unlike most solid tumors, HCC typically arises in the context of underlying liver disease—cirrhosis from hepatitis B, hepatitis C, alcoholic liver disease, or non-alcoholic steatohepatitis (NASH/MAFLD)—creating a dual management challenge: treating the tumor while preserving residual liver function in an already-compromised organ.
Treatment is stage-dependent and liver-function-dependent. Early-stage disease (BCLC 0/A) is addressed with surgical resection, radiofrequency or microwave ablation, or liver transplantation. Intermediate-stage disease (BCLC B) receives transarterial chemoembolization (TACE) or selective internal radiation therapy (SIRT/Y90). Advanced-stage disease (BCLC C) is managed systemically with sorafenib or lenvatinib as first-line, and atezolizumab+bevacizumab (IMbrave150) has become the preferred first-line for patients without contraindications. Second-line options include cabozantinib, regorafenib, and ramucirumab (AFP ≥400 ng/mL).
Digital health platforms supporting HCC programs must coordinate AFP surveillance, contrast-enhanced MRI/CT imaging intervals, liver function trajectory monitoring (Child-Pugh, MELD scores), portal hypertension management, and transplant listing workflows. This guide covers how to build and monitor those platforms for reliability and clinical impact in 2026.
Care Technology Landscape
AFP Trending and Biomarker Surveillance Platforms
Alpha-fetoprotein (AFP) remains the primary serum biomarker for HCC surveillance and treatment response assessment, though its sensitivity is modest (60–70%) for early disease. AFP-L3 fraction and des-gamma-carboxyprothrombin (DCP/PIVKA-II) provide complementary diagnostic signals. Platforms must track AFP trajectories longitudinally, calculate velocity (rate of rise), and surface deviations from baseline to hepatologists at scheduled follow-up intervals and between visits when results return outside expected ranges.
Contrast-Enhanced MRI/CT Imaging Surveillance Schedulers
High-risk cirrhotic patients require semi-annual surveillance imaging per AASLD/EASL guidelines. HCC diagnosis is typically established by imaging criteria alone (LI-RADS 4/5 on contrast-enhanced CT or MRI) without biopsy. Imaging platforms must schedule and track surveillance intervals, coordinate with radiology for protocol-compliant contrast-enhanced sequences, capture LI-RADS categorizations from structured radiology reports, and alert care teams when imaging is overdue or when an LI-RADS upgrade warrants expedited tumor board review.
Child-Pugh / MELD Score Progression Monitoring
Liver synthetic function is tracked via Child-Pugh score (bilirubin, INR, albumin, ascites, encephalopathy) and MELD score (bilirubin, INR, creatinine), which together determine surgical candidacy, transplant listing priority, and systemic therapy eligibility (bevacizumab is contraindicated in decompensated cirrhosis). Platforms must calculate and trend these scores at each relevant lab update, alert on decompensation events (score progression, new ascites, encephalopathy episodes), and maintain transplant listing readiness assessments.
Portal Hypertension Management Platforms
Cirrhosis-related portal hypertension produces life-threatening complications: variceal hemorrhage, spontaneous bacterial peritonitis (SBP), and hepatorenal syndrome. Digital health platforms coordinate esophagogastroduodenoscopy (EGD) surveillance for varices, track non-selective beta-blocker therapy adherence, monitor diagnostic paracentesis results for SBP, and log hepatic encephalopathy episodes. Integration with hepatology care coordination ensures that HCC treatment does not inadvertently worsen portal hypertension management.
Transplant Listing Status Monitoring
Liver transplantation is the only potentially curative option for patients within Milan or UCSF criteria who are not surgical resection candidates due to underlying liver disease. Transplant listing platforms track MELD score dynamics, Milan/UCSF criteria compliance over time, bridging therapy outcomes (TACE, ablation), HCC exception points requests to UNOS/OPTN, and waitlist time management. Automated alerts flag patients approaching criteria boundaries (tumor growth approaching size limits, MELD changes) requiring urgent re-evaluation.
Key Monitoring Metrics
Clinical Workflow Metrics
| Metric | Target | Alert Threshold | |---|---|---| | Surveillance imaging interval adherence (high-risk cirrhosis) | >90% within 6-month window | <80% | | AFP result turnaround (routine surveillance) | ≤2 days from blood draw | >5 days | | LI-RADS 4/5 upgrade to tumor board presentation | ≤7 days | >14 days | | MELD score recalculation at each lab update | <1 hour from result | >4 hours | | Transplant listing re-evaluation when Milan criteria approached | ≤48 hours | >72 hours |
Platform Infrastructure Metrics
| Metric | Target | Alert Threshold | |---|---|---| | AFP/lab result LIS ingestion latency | <5 minutes | >15 minutes | | LI-RADS structured report extraction accuracy | >98% | <95% | | MELD auto-calculation success rate | >99.5% | <98% | | Imaging scheduler uptime | 99.9% | <99.5% | | Transplant listing API integration uptime | 99.9% | <99.5% |
Patient Safety Metrics
| Metric | Target | Alert Threshold | |---|---|---| | Cirrhotic patients overdue for surveillance imaging (>7 months) | <5% of active cohort | >10% | | Child-Pugh C patients on bevacizumab-containing regimen without documented contraindication review | 0 | Any | | Transplant candidates with AFP >1000 ng/mL without documented treatment intensification plan | 0 | Any | | Variceal hemorrhage patients without EGD within 12 hours | 0 | Any |
Platform Setup
1. AFP Velocity and Trend Analytics
Configure longitudinal AFP tracking with automated velocity calculation (ng/mL/month) at each new result. Set individualized baseline ranges for each patient (AFP can be chronically elevated in cirrhosis without HCC) and alert on velocity exceeding thresholds or absolute values crossing clinical decision points.
# AFP surveillance configuration
afp_monitoring:
baseline_window_days: 90
velocity_alert:
threshold_ngml_per_month: 15
alert_to: [hepatologist, tumor_board_coordinator]
absolute_thresholds:
200: advisory_review
400: ramucirumab_eligibility_flag
1000: urgent_multidisciplinary_review
companion_biomarkers:
- AFP_L3_fraction
- DCP_PIVKA_II
correlation_check: true
2. LI-RADS Structured Report Integration
Integrate with the radiology information system (RIS) to automatically extract LI-RADS category from structured CT/MRI reports. Use NLP parsing or structured template enforcement to capture LI-RADS category, tumor size, number of lesions, and vascular invasion status from each surveillance imaging report.
# LI-RADS imaging monitoring
imaging_surveillance:
modalities: [CT_triphasic, MRI_hepatobiliary_protocol]
interval_months: 6
lirads_extraction:
method: structured_template
fallback: nlp_parser
fields_required: [lirads_category, max_lesion_size_cm, lesion_count, vascular_invasion]
alert_rules:
- lirads_category: [LR4, LR5]
action: [tumor_board_referral, hepatologist_notification]
urgency: 7_days
- lirads_category: LR_M
action: [oncology_referral, biopsy_consideration_task]
urgency: 7_days
overdue_alerts:
last_imaging_days_threshold: 210
notify: [care_coordinator, hepatologist]
3. Automated Child-Pugh and MELD Calculation
Configure real-time score calculation triggered by any lab result update affecting the components. Child-Pugh requires manual clinical inputs (ascites grade, encephalopathy grade) that should be prompted at each hepatology encounter. Build a MELD calculator that ingests bilirubin, INR, and creatinine automatically from the LIS feed and recalculates within minutes of result availability.
# Liver function score monitoring
liver_function_scores:
meld_components:
- bilirubin_total
- INR
- creatinine
auto_recalculate_on_new_lab: true
child_pugh_components:
lab_auto: [bilirubin_total, albumin, INR]
clinical_prompted: [ascites_grade, encephalopathy_grade]
decompensation_alerts:
meld_increase_7d: 5
child_pugh_class_change: A_to_B_or_B_to_C
new_ascites_grade_3: true
new_encephalopathy_grade_2_plus: true
transplant_integration:
auto_update_unos_meld: true
listing_review_trigger_meld: 15
4. Portal Hypertension Complication Surveillance
Implement a structured complications tracker for cirrhosis management, capturing EGD-confirmed variceal grade, beta-blocker therapy status, SBP episodes with culture results, hepatic encephalopathy episodes, and hepatorenal syndrome events. Integrate with the HCC treatment planning system to ensure portal hypertension status is reviewed before TACE or systemic therapy initiation.
Alerting Strategies
Tiered Alert Classification
P1 – Immediate (page on-call hepatologist):
- Variceal hemorrhage admission (GI bleed + known varices) without endoscopy order within 2 hours
- AFP velocity >50 ng/mL/month in a transplant candidate within Milan criteria
- New LI-RADS 5 finding in a patient not currently under active HCC treatment
- Child-Pugh C patient initiated on bevacizumab without documented review
P2 – Urgent (notify within 4 hours):
- MELD increase ≥5 points within 7 days (acute decompensation signal)
- Surveillance imaging overdue >30 days for any high-risk cirrhotic patient
- AFP >1,000 ng/mL in patient not yet discussed at tumor board
- Transplant candidate approaching Milan criteria boundary (tumor growth within 10% of size limit)
P3 – Advisory (daily digest):
- Overall surveillance imaging adherence below monthly target
- LI-RADS structured report extraction failure rate above 2%
- MELD auto-calculation delays averaging >15 minutes
Cross-Domain Alert Coordination
HCC care inherently spans hepatology, oncology, interventional radiology, and transplant surgery. Configure alert routing to deliver relevant notifications to each team without overwhelming any single service:
alert_routing_matrix:
AFP_elevation:
hepatology: P2_notify
oncology: P2_notify
transplant: P2_if_transplant_candidate
LI_RADS_upgrade:
hepatology: P1_notify
oncology: P2_notify
ir_radiology: P2_tace_eligibility_review
MELD_decompensation:
hepatology: P1_notify
transplant: P1_if_listed
oncology: P2_systemic_therapy_review
variceal_hemorrhage:
hepatology: P1_page
transplant: P2_listing_impact_review
oncology: P2_treatment_hold_consideration
Surveillance Cohort Management
For high-risk cirrhosis populations (often hundreds of patients per hepatology practice), a population-level surveillance dashboard is essential alongside individual patient alerts. Build a dashboard view showing:
- Patients due for imaging within the next 30 days
- Patients overdue by >1 month, >3 months, and >6 months
- AFP trends by patient cluster (stable, rising, declining on treatment)
- Active HCC treatment stage distribution (bridging therapy, systemic, transplant wait)
Conclusion
HCC care tech platforms operate at the intersection of oncology and advanced liver disease, requiring monitoring capabilities that span tumor biology, hepatic function, portal hypertension complications, locoregional procedure outcomes, and transplant eligibility. The platforms that serve this patient population must be exceptionally reliable—decompensation events and tumor progression can occur rapidly, and surveillance gaps directly translate to missed curative opportunities.
Key priorities for 2026 HCC platform operations:
- Semi-annual imaging surveillance adherence to catch early-stage HCC when curative options are available
- AFP velocity analytics to provide early warning of tumor growth between imaging intervals
- Real-time MELD tracking to guide transplant listing priority and treatment eligibility decisions
- Cross-team alert coordination ensuring that hepatology, oncology, IR, and transplant surgery all receive timely, relevant notifications without duplication or gaps
Vigilmon's monitoring infrastructure—synthetic checks, uptime dashboards, tiered alert management, and multi-channel routing—provides the observability backbone that HCC care platforms need to function without gaps in a disease where timing is everything.