Search

518 Background: Final results from the real-world REFINE study confirmed the safety and effectiveness of regorafenib (REG) in a broad population of patients (pts) with uHCC that was consistent with the results from RESORCE, in which REG prolonged overall survival (OS) vs placebo in pts with uHCC who progressed on prior sorafenib. Here we evaluate baseline characteristics and outcomes according to the initial REG dose. Methods: REFINE (NCT03289273), an international, prospective, multicenter study, enrolled pts with uHCC for whom a decision to treat with REG was made by the treating physician prior to enrollment according to the local health authority approved label (160 mg/day, 3 weeks on/1 week off). The primary objective was safety (NCI-CTCAE v4.03); secondary endpoints included OS and progression-free survival (PFS). Results: Of 1005 pts eligible for analysis, 47%, 11%, 40%, and 3% initiated REG at 160 mg, 120 mg, 80 mg, and 40 mg, respectively. A higher proportion of pts from Asia initiated REG at 160 mg (51%) vs non-Asian countries (42%), and 67% were Child–Pugh A in Asia vs 54% in non-Asian countries at study entry. Pts initiating REG at 160 mg were more likely to be Child–Pugh A and albumin–bilirubin (ALBI) grade 1 vs pts initiating REG at

433 Background: Regorafenib improved OS versus placebo in patients with uHCC who progressed on prior sorafenib in the RESORCE trial (Bruix J, 2017). An interim analysis of the observational REFINE study supported the safety and effectiveness of regorafenib in patients with uHCC in real-world clinical practice (Lim HY, 2021). Here we present an exploratory analysis of OS in REFINE by prior treatment. Methods: REFINE is a multicenter study that enrolled patients with uHCC for whom a decision to treat with regorafenib is made by the treating physician prior to enrollment according to the local health authority approved label. The primary objective of this study is safety, including the incidences of treatment-emergent adverse events (TEAEs) and dose modifications due to TEAEs (NCI-CTCAE v4.03). The secondary endpoints include OS, progression-free survival, and treatment duration. Results: Of the 1,031 patients enrolled, 1,008 were evaluable for interim analysis. At baseline, median age was 66 years (range 21–94); 62% of patients were Barcelona Clinic Liver Cancer stage C; 62% had Child–Pugh A disease; and 83% had an Eastern Cooperative Oncology Group performance status of 0 or 1. In total, 99% of patients received prior treatment: 96% had prior sorafenib, 9% had ≥1 prior immunotherapy (most common: nivolumab [50%] and pembrolizumab [21%]), and 6% had a multikinase inhibitor other than sorafenib (lenvatinib [62%]). The majority of patients experienced TEAEs (91%) and drug-related TEAEs (73%). Median OS was 12.9 months (95% confidence interval [CI] 11.4, 14.6). Subgroup analyses of OS by prior treatment are shown (Table). Conclusions: REFINE reflects the changing treatment landscape and supports the safety and effectiveness of regorafenib in real-world patients with uHCC who had prior systemic treatment other than sorafenib, including immunotherapy. This interim analysis suggests that patients who received regorafenib in the second line had longer OS than patients who received regorafenib in the third line and beyond. Clinical trial information: NCT03289273. [Table: see text]

412 Background: The standard treatment for intermediate stage hepatocellular carcinoma (HCC) has been transarterial chemoembolization (TACE); however, it might not be suitable for patients (pts) beyond the up-to-7 criteria because of its insufficient efficacy and decline of liver function. Conversely, lenvatinib has shown a good response rate in the phase III trial REFLECT and was considered promising as an initial treatment for patients with BCLC stage B2 HCC. Methods: We conducted a multicenter prospective observational study to evaluate the efficacy of lenvatinib as an initial treatment for pts aged 20 years or older, diagnosed with BCLC stage B2 HCC, naive to TACE, and with preserved organ function. The primary endpoint was a 1-year survival rate, with a threshold of 60% and an expected survival rate of 78% based on previous reports of TACE. With a one-sided alpha error of 5% and a power of 70%, 25 pts were recruited over an enrollment period of 2 years and a follow-up period of 10 months. Results: Enrollment began in June 2018 and was closed with 31 eligible pts enrolled by June 2020. Data cut-off was done in April 2021. Patient characteristics are as follows: the median age was 77 years (range: 57–86); 94% of the pts were male; the etiology of chronic liver disease was hepatitis B virus, hepatitis C virus, alcohol abuse, and others in 3, 6, 9, and 10 pts, respectively; 20 and 11 pts had a Child-Pugh score of 5 and 6 points, respectively; the median maximum tumor diameter was 36 mm (range: 10–135); the number of pts with a tumor number of ≥10 was 8; and the median AFP level was 52.1 ng/mL (range: 2.4–49800). The 1-year survival rate was 71.0% (90% confidence interval [CI]: 68.4–73.6) and the primary endpoint was met. The median overall and progression-free survival was 17.0 months (95% CI, 15.3–19.2) and 10.4 months (95% CI, 6.6–13.8), respectively; the 2-year survival rate was 32.3%. The objective response rate according to RECIT1.1 and mRECIST was 22.6% and 70.0%, respectively. The common adverse events (AEs) were fatigue (68%), hypertension (65%), and anorexia (61%) in any grade, AST increased (23%), ALT increased (16%), and proteinuria (13%) in grade 3 or worse. Treatment interruption and dose reduction was required in 61% and 81% of the pts, respectively. The median relative dose intensity was 61.8% (range: 7.5–100). At the time of analysis, out of 24 pts, 17 (55%), nine (29%), and two (6%) discontinued lenvatinib due to disease progression, AEs, and conversion to curative treatments, respectively. One patient who showed partial response suddenly died during treatment due to unknown reasons. Systemic chemotherapy, TACE, and transarterial infusion therapy was administered in 11 (38%), six (21%) and one (3%) as a post study treatment, respectively. Conclusions: Our study indicates that initial treatment using lenvatinib prolongs the survival of pts with BCLC stage B2 HCC.

e16680 Background: REG is approved for the treatment of pts with HCC who previously received SOR, based on results of the phase 3 RESORCE trial. The REFINE study was designed to evaluate the safety and effectiveness of REG in pts with HCC in real-world practice. Methods: This prospective, observational study aims to recruit 1000 pts with unresectable HCC for whom a decision to treat with REG was made by the treating physician prior to enrollment according to the local health authority approved label. The primary endpoint is the incidence of treatment-emergent adverse events (TEAEs) and dose modifications due to TEAEs (NCI-CTCAE v4.03). Secondary endpoints include overall survival (OS) and progression-free survival (investigator assessed). This interim analysis includes the first 500 pts on study for ≥4 months. Results: Of 500 pts enrolled and observed for ≥4 months, 498 received REG and were evaluable (data cut-off: November 11, 2019). Pts were Child-Pugh A 67%, B 11%, C 1%, and missing/not evaluable 21%. Most pts (98%; n = 490) had received prior systemic therapy; 97% (n = 482) had received prior SOR. REG was second line treatment in 81% of pts (n = 403), third line or higher in 17% (n = 87), and first line in 2% (n = 8). Of the 403 pts who received REG second line, 398 (99%) received prior SOR. Among the 482 pts who received SOR in any prior line, the median duration of prior SOR was 4.8 months (interquartile range 2.5–9.6), 45% (n = 216) had a last daily SOR dose of 800 mg, and 8% (n = 40) had a TEAE leading to SOR discontinuation (SOR-intolerant). OS by treatment line is shown in the Table. Among all treated pts, the most frequent TEAEs (any grade) were hand–foot skin reaction (HFSR; 30%), diarrhea (21%), fatigue (16%), and decreased appetite (14%). In SOR-intolerant pts (n = 40), the most frequent TEAEs (any grade) were diarrhea (25%), HFSR (20%), abdominal pain (15%), and decreased appetite (13%). Conclusions: In this interim analysis of REFINE, most pts received REG second line after SOR versus other lines of therapy. Median OS in this subgroup was longer than OS in RESORCE, but the proportion of censored pts was high. The most common TEAEs were similar to those in RESORCE in both the overall cohort and in SOR-intolerant pts. Clinical trial information: NCT03289273 . [Table: see text]

327 Background: Pts with aHCC and CPB liver status are often excluded from clinical trials of novel therapies due to their poor prognosis (Greten British J Cancer 2005). Historical overall survival (OS) for these pts when treated with sorafenib (SOR) has ranged ≈3–5 mo in retrospective or descriptive studies (Abou-Alfa Gastrointest Cancer Res 2011; Da Fonseca Mol Clin Oncol 2015; Pressiani Ann Oncol 2013; Chiu Cancer 2012); thus, novel treatment options are needed for these pts. The PD-1 inhibitor NIVO is approved in the US, Canada, and elsewhere, most recently Australia, for SOR-treated pts with aHCC based on results from CheckMate-040 (NCT01658878) (El-Khoueiry Lancet 2017). Here we report data from the CPB cohort of CheckMate-040, the first prospective study of immunotherapy in this pt group. Methods: Pts with CPB (B7–B8) aHCC who were SOR-naïve (n = 25) or -experienced (n = 24) received NIVO 240 mg IV for 30 min Q2W until unacceptable toxicity or disease progression. Primary endpoints were objective response rate (ORR) (investigator assessed [INV], RECIST v1.1) and duration of response (DOR). Safety was assessed in all treated pts using NCI CTCAE v4.0. Results: Of 49 analyzed pts, 28 (57.1%) had vascular invasion or extrahepatic spread. During a follow-up range of 6–18 mo, INV ORR was 10.2% with 5 pts responding; disease control rate (DCR) was 55.1%. Median (m) time to response was 2.7 mo and mDOR was 9.9 mo; 2 pts had ongoing responses at data cutoff. The mOS was 7.6 mo (mOS follow-up was 7.4 mo); mOS in SOR-naïve and -treated pts was 9.8 and 7.3 mo, respectively. Treatment-related adverse events (TRAEs) were reported in 25 (51%) pts; 4 (8.2%) pts had select hepatic TRAEs. TRAEs led to discontinuation in 2 pts (4.1%). NIVO safety profile in these pts appeared comparable to cohorts of pts with CPA aHCC. Comparison data for pts with CPA aHCC and extended follow-up for pts with CPB aHCC will be presented. Conclusions: Encouraging DCR and durable responses were observed in pts with CPB aHCC treated with NIVO. AEs were manageable and did not lead to higher discontinuation compared with pts with CPA aHCC. NIVO showed promising efficacy and tolerability compared with historical data, supporting further investigation. Clinical trial information: NCT01658878.

4018 Background: An unmet medical need persists for patients (pts) with sorafenib-refractory advanced hepatocellular carcinoma (HCC). This study was conducted to evaluate the efficacy and safety of...

237 Background: HCC is a highly vascularized tumor, and inhibition of angiogenesis by sorafenib (Sor)—a VEGFR and PDGFR inhibitor—delays tumor progression. However, angiogenic escape from Sor may result from activation of the FGFR pathway, which also plays an important role in angiogenesis. Dovitinib (Dov) inhibits FGFR as well as VEGFR and PDGFR. Here, we study frontline Dov vs Sor in pts with advanced HCC. Methods: Eligible pts in this open-label study had ≥ 1 measurable lesion at baseline. All pts were ineligible for or had disease progression after surgical and/or locoregional therapies. Prior systemic HCC therapy was not allowed. Pts were randomized to receive Dov (500 mg/day, 5 days on/2 days off) or Sor (400 mg twice daily) until disease progression, unacceptable toxicity, or death. No treatment crossover was allowed. The primary endpoint was overall survival (OS). The key secondary endpoint was time to tumor progression (TTP) by local investigator’s assessment (per RECIST v1.1). Results: Pts received Dov (n = 82) or Sor (n = 83). Most pts—43 (52.4%) in the Dov arm and 60 (72.3%) in the Sor arm—discontinued treatment due to progressive disease. Median OS (95% CI) was 34.6 wk (28.6-39.4 wk) for Dov and 36.7 wk (23.3-49.3 wk) for Sor; Kaplan-Meier HR, 1.27; 95% CI, 0.89-1.80. Median TTP (95% CI) was 17.6 wk (12.3-18.4 wk) and 17.9 wk (12.3-18.9 wk), respectively. In pts who received ≥ 1 dose of study drug, median duration of exposure was 2.5 mo for Dov (n = 79) and 3.2 mo for Sor (n = 83). The most common adverse events, regardless of cause, were diarrhea (62.0%), decreased appetite (43.0%), nausea and vomiting (40.5% each), fatigue (35.4%), rash (34.2%), and pyrexia (30.4%) in the Dov arm and palmar-plantar erythrodysesthesia syndrome (66.3%), diarrhea (42.2%), and decreased appetite (31.3%) in the Sor arm. VEGFR1 and HGF baseline plasma levels appear to be associated with OS only in the Dov arm. Median OS in the lower biomarker group for both VEGFR1 and HGF was 11 mo while it was 5.6 and 5.9 mo for VEGFR1 and HGF, respectively, in the higher biomarker group. Conclusions: Activity of Dov was not greater than that of Sor in frontline HCC. The safety profile was similar to that observed in other single-agent Dov trials. Clinical trial information: NCT01232296.

15673 Background: Recent studies have shown the prognosis and the prognostic factors for hepatocellular carcinoma (HCC) after initial transcatheter arterial chemoembolization (TACE), however, no st...