#4.
Background Multiple regimens are available as first-line systemic therapy for advanced hepatocellular carcinoma (HCC).1 Atezolizumab plus bevacizumab (Atezo+Bev) has demonstrated favorable efficacy and safety profiles and has become the most commonly used first-line treatment for HCC.2 However, we previously reported that the efficacy of Atezo+Bev is limited in Japanese patients with a CRAFITY score of 2 (i.e., CRP ≥ 1 mg/dL and AFP ≥ 100 ng/mL),3 and that lenvatinib (LEN) provided better progression-free survival (PFS) than Atezo+Bev in this population.4 To validate these findings, here we conducted an international collaborative study comparing the efficacy of Atezo+Bev and LEN stratified by the baseline CRAFITY score. 1. Vogel A, et al. Ann Oncol. 2025;36:491–506. 2. Asaoka Y, et al. Liver Cancer. 2025 (Epub ahead of print). 3. Ueno M, et al. J Gastroenterol. 2024;59:1107–1118. 4. Ueno M, et al. Hepatol Res. 2026;56:100–110.
#5.
Our concept of personalized treatment for HCC Advanced HCC Child-Pugh A, ECOG PS 0-1 Is patient a candidate for immunotherapy? Yes No ASCO, ESMO, EASL, AASLD, AGA, NCCN High risk of GI bleeding? No Atezo＋Bev CRAFITY score 2 ASCO, ESMO, EASL, AASLD, AGA Yes Autoimmune disorder? Liver transplant? AASLD Durva＋Treme LEN Ueno M, et al. JSMO 2025. PS4-1.
#6.
Aims of the present study To validate our previous finding that LEN has better efficacy than Atezo+Bev in patients with a CRAFITY score of 2.
#7.
Methods Patients｜Among 1449 HCC patients treated with 1st-line Atezo+Bev or LEN between September 2017 and March 2024 across 11 hospitals in Japan and 15 hospitals in Taiwan, 994 patients with baseline CRP and AFP evaluation were enrolled Data collection｜Baseline patients’ characteristics and laboratory data Follow-up data up to July 2024 were collected Calculation of CRAFITY score｜AFP ≥100 ng/mL and CRP ≥1 mg/dL are each assigned 1 point Endpoints｜PFS (RECIST v1.1) Overall survival (OS) Objective response rate, Disease control rate Treatment-related adverse events (TRAEs)
#8.
CRAFITY-0 | PFS and OS PFS Regimen median PFS OS Regimen median PFS Atezo+Bev 10.8 months Atezo+Bev 21.1 months LEN 9.4 months LEN 29.3 months HR = 1.09 [95% CI, 0.82–1.44] (p = 0.55) HR = 0.72 [95% CI, 0.50–1.03] (p = 0.07)
#9.
CRAFITY-1 | PFS and OS PFS Regimen median PFS OS Regimen median PFS Atezo+Bev 6.0 months Atezo+Bev 12.7 months LEN 6.1 months LEN 15.1 months HR = 0.99 [95% CI, 0.76–1.29] (p = 0.94) HR = 0.82 [95% CI, 0.61–1.09] (p = 0.17)
#10.
CRAFITY-2 | Baseline characteristics Atezo+Bev LEN p value (n = 62) (n = 155) 67.5 [58.3–75] 70 [61–77] 0.25 53 / 9 113 / 42 0.07 27 / 25 / 7 59 / 52 / 26 0.45 34 / 28 84 / 71 1.00 16 / 19 / 14 / 6 / 4 44 / 57 / 21 / 13 / 8 0.27 35 / 24 101 / 42 0.14 -2.00 [-2.36 to -1.72] -2.15 [-2.53 to -1.77] 0.07 BCLC stage ( A / B / C ) 12 / 73 / 68 23 / 91 / 76 0.25 Macrovascular invasion 26 (17.0%) 25 (13.2%) 0.40 Extrahepatic spreading 44 (28.8%) 54 (28.4%) 1.00 Age Sex ( Male / Female ) PS ( 0 / 1 / ≥2 ) Etiology ( Viral / Non-viral ) Child–Pugh score ( 5 / 6 / 7 / 8 / ≥9 ) Child–Pugh class ( A / B ) ALBI score Expressed as number (%) or median [IQR]
#11.
CRAFITY-2 | PFS and OS PFS Regimen median PFS OS Regimen median PFS Atezo+Bev 2.3 months Atezo+Bev 5.7 months LEN 4.4 months LEN 9.1 months HR = 0.66 [95% CI, 0.47–0.93] (p = 0.01) HR = 0.90 [95% CI, 0.61–1.32] (p = 0.59)
#12.
CRAFITY-2 | PFS stratified by country Japan Regimen median PFS Taiwan Regimen median PFS Atezo+Bev 1.8 months Atezo+Bev 2.3 months LEN 4.9 months LEN 4.3 months HR = 0.61 [95% CI, 0.31–1.19] (p = 0.15) HR = 0.65 [95% CI, 0.44–0.97] (p = 0.04)
#13.
Conclusions LEN provided significantly longer PFS than Atezo+Bev in patients with a baseline CRAFITY score of 2 in this international collaborative analysis. In patients with a CRAFITY score of 0 or 1, the two regimens showed broadly similar efficacy. The CRAFITY score is a promising biomarker for identifying patients who may derive greater benefit from LEN than from Atezo+Bev.
