Peg-interferon add-on clears HBsAg in 38% - if the starting antigen is below 100 IU/mL

Design

post-hoc individual participant data meta-analysis of 8 prospective trials

Population

581 patients with chronic hepatitis B on nucleos(t)ide analogues receiving peg-interferon add-on

Primary outcome

HBsAg loss 6-12 months after end of peg-interferon

Effect

8.6% overall; 37.7% below 100 IU/mL at start, 9.8% at 100-1000, 2.3% above 1000 (p<0.001)

PROSPER pooled individual participant data from eight trials in which patients with chronic hepatitis B established on nucleos(t)ide analogues received peg-interferon as an add-on: 581 patients, 44% HBeAg positive, mean HBsAg 3.03 log10 IU/mL, with 48 weeks of peg-IFN planned in 85%. The endpoint was HBsAg loss 6 to 12 months after peg-IFN stopped.

Overall, 8.6% achieved it. Stratified by the antigen level at the start of peg-IFN, the figure separates into three quite different treatments: 37.7% below 100 IU/mL, 9.8% between 100 and 1000, and 2.3% above 1000 (p<0.001). The gradient held across ethnicity - 39.3%, 9.2% and 2.2% in Asian patients against 30.0%, 8.7% and 3.6% in Caucasian patients.

For patients starting above 1000 IU/mL the paper suggests a different purpose. Forty-eight weeks of peg-IFN moved 47.5% of them below 1000 IU/mL and 16.3% below 100 - into the range where the agents now in development are most likely to work. That is a case for peg-IFN as a lead-in within a trial pathway, not as a cure attempt in routine care, and it rests on a surrogate.

• Measure quantitative HBsAg before offering peg-interferon add-on; the decision depends on it
• Quote about 38% functional cure below 100 IU/mL and about 2% above 1000 - not the 8.6% average
• Weigh 48 weeks of interferon toxicity against a 2% chance honestly in high-antigen patients
• Continue the nucleos(t)ide analogue throughout - this is add-on therapy
• Consider trial referral rather than routine peg-IFN where the antigen is high and a novel agent is in reach

Why it matters

It turns peg-interferon add-on from a treatment offered on general grounds into one with a selection criterion you can already measure.

Don't overread it

HBsAg loss is a surrogate endpoint - nothing here demonstrates fewer cases of cirrhosis, liver cancer or death.

The statistics, in plain English

The 8.6% overall figure averages across three strata whose results differ more than tenfold, which makes it misleading for any individual - this is a case where the pooled number obscures the finding rather than summarising it. The stratification was applied post hoc, so the thresholds describe these data rather than being validated cut-offs. And HBsAg loss is a surrogate: the trials were not powered for cirrhosis, hepatocellular carcinoma or death, which is what patients are actually being treated to avoid.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.