Immunotherapy added to TACE delays progression in liver cancer, and a blood test outperforms surveillance ultrasound

The edition in brief

EMERALD-3 randomised 760 people with Child-Pugh A hepatocellular carcinoma suitable for transarterial chemoembolisation to durvalumab plus tremelimumab (STRIDE) with lenvatinib and TACE, STRIDE with TACE, or TACE alone. Adding STRIDE and lenvatinib lengthened median progression-free survival from 9.8 to 13.0 months (HR 0.70, 95% CI 0.57-0.86). Overall survival was 39.5 versus 34.7 months, but the interval crosses 1.0 (HR 0.84, 0.65-1.09) and the analysis is immature. Serious adverse events rose from 23% to 64%, with seven treatment-related deaths in the triplet arm. A retrospective cohort of 22,537 veterans with alpha-1 antitrypsin genotyping found liver risk rising with Z-allele burden rather than starting at Pi*ZZ. Adjusted hazards for major adverse liver outcomes were 1.25 for Pi*MZ, 1.51 for Pi*SZ and 1.80 for Pi*ZZ against Pi*MM, with five-year probabilities of 5.5%, 5.3% and 8.1% versus 3.5%. Heterozygotes carried more decompensation, transplantation and liver-related death, though not more hepatocellular carcinoma. A meta-analysis of 126 studies and 86 million participants set liver cancer risk beside the better-known risk of lung cancer from smoking. Against an odds ratio of 6.0 for current smoking and lung cancer, hepatitis B carried 15.2, hepatitis C 12.8 and hepatitis D 23.5, while MASLD (2.6) and alcohol (2.4) sat lower. Regional differences were not significant. The FDA acted on a supplemental application for the adalimumab biosimilar Yusimry; the record gives the action, not what changed. In a blinded prospective study of 1,268 patients with cirrhosis and MRI as reference, a multianalyte cell-free DNA blood test detected 47.8% of hepatocellular carcinomas against ultrasound's 28.3%, and 28.6% of lesions of 2 cm or less where ultrasound found none - at the cost of lower specificity, 87.6% versus 93.9%.