Every meta-analysis comparing treatments for hepatocellular carcinoma, mapped and kept current.
GRAIL-HCC charts where the comparative evidence on HCC treatments concentrates and where it is missing, across the seven tiers of the multiparametric therapeutic hierarchy. The map is updated annually, moving to six-monthly updates as the programme matures, and is free to use.
The map today
What it shows
- A few questions absorb much of the literature: resection versus ablation, chemoembolization plus systemic therapy, immunotherapy- versus kinase-inhibitor-based regimens.
- The surgical apex of the hierarchy is the least synthesised. Resection has never been compared with systemic therapy in a dedicated meta-analysis.
- Only 11% of reviews rated certainty of evidence with GRADE, and 7% reported quality of life.
How to read it
Each bubble counts meta-analyses, not patients or effects. Colour shows the share of reviews based only on randomised trials. Volume and design are not quality: methodological appraisal with AMSTAR 2 is in progress.
Click any bubble or bar in the map to list the reviews behind it, with links to PubMed.
A programme in three phases
1 · Evidence and gap map
All comparative meta-analyses of HCC treatments since 2010, classified by comparison class and hierarchy tier, updated annually and then every six months.
2 · Umbrella reviews
For each prioritised comparison: overlap of primary studies, index meta-analyses chosen by recency, size and AMSTAR 2, pooled effects, and certainty rated with GRADE.
3 · Decision support
Evidence snapshots for the multidisciplinary tumour board, built on GRADE Evidence-to-Decision frameworks for the individual patient.
Who we are
GRAIL-HCC is led from the Department of Surgical, Oncological and Gastroenterological Sciences (DISCOG) of the University of Padua. More than 90 hepatologists, surgeons, interventional radiologists, radiation oncologists, medical oncologists and methodologists from over 30 Italian centres take part.
Meet the working groups →Use and cite
The map and its record-level dataset are open under CC BY 4.0. Please cite the version you used.
How to cite →