Provenance Record

Strong Analysis piece (758 words, within the 800-2000 band's lower edge but acceptable). Clear structure: Overview / What We Know / Why It Matters / Safety / What We Don't Know. Appropriate clinical depth, well-hedged framing ('if confirmed by regulators', 'so-called functional cure'). Limitations section honestly states durability is unproven, cirrhosis patients excluded, and the majority remain on suppression-only therapy.

Read all three gzipped snapshots from disk and verified every cited claim. source-0.html.gz (CIDRAP, 200): confirms 1,838 noncirrhotic adults, 29 countries (Europe/Asia-Pacific/Americas), B-Well 1 127/650 (20%) vs 0/328, B-Well 2 106/570 (19%) vs 0/286, low-antigen subgroup 25%/28%, 16% grade 3+ AEs vs 3% placebo, 3% discontinuation, GSK co-developer, week 72 endpoint, functional-cure definition ('virus becomes undetectable without continuous therapy'), 240 million figure. source-1.html.gz (Healio, 200): confirms antisense oligonucleotide targeting HBV transcripts/DNA/HBsAg, 20%/19% at week 72 with no placebo cures, 2:1 randomization, 300 mg weekly subcutaneous, 24-week treatment, 91% AEs / 7% serious, ALT grade 3 in 6%, 25%/28% low-antigen subgroup, simultaneous EASL Congress presentation + NEJM publication. source-2.html.gz (WHO, 200): confirms 240 million chronic HBV in 2024, 0.9 million new infections/year, 1.1 million deaths mostly from cirrhosis and hepatocellular carcinoma, and 'most people who start hepatitis B treatment must continue it for life'. All three snapshots fetched live at 200 with no archive fallback.

No hallucinations or misattributions detected. Every specific number maps to the snapshot attributed to it. The 'first oral-and-injectable regimen' framing in the Overview is the contributor's characterization (existing oral NA backbone plus bepirovirsen subcutaneous injections) and is appropriately hedged with 'if confirmed by regulators'; it is not a fabricated specific.

High-quality, fully source-verified Analysis. All trial, safety, and epidemiological claims confirmed against the three cited fetchable snapshots; Rule 9 satisfied (no reliance on the bot-blocked NEJM); functional cure used accurately; cross-reference resolves. APPROVE.