Search

506 Background: In LEAP-002 (NCT03713593), len + pembro achieved a median OS of 21.2 mo vs 19.0 mo with len + pbo with manageable safety in both arms in 1L aHCC, although the significance thresholds for OS and PFS were not met. We present here the results of prespecified exploratory patient (pt)-reported outcomes (PROs). Methods: 794 pts were randomized. PROs were assessed by EORTC QLQ-C30, EuroQol-5D5L (EQ5D-5L), and EORTC QLQ-HCC18 questionnaires. Analysis population for HRQoL endpoints included pts who received ≥1 dose of study Tx and completed ≥1 HRQoL assessment. Least squares mean (LSM) score changes from baseline (BL) to wk 27 were compared using a constrained longitudinal data analysis model (covariates: Tx, time, Tx by time interaction, and stratification factors). Kaplan-Meier method was used to estimate time to deterioration (TTD, time to 1st onset of ≥10 [out of 100] deterioration from BL in a given scale or subscale/confirmed by a 2nd adjacent ≥10 deterioration from BL) for EORTC QLQ-C30 global health status (GHS)/QoL, physical functioning (PF), and EORTC QLQ-HCC18 abdominal swelling, pain and fatigue, as reported by pts. Stratified Cox proportional hazards model was used to assess the magnitude of the Tx difference (HR) between arms in TTD. Results: Compliance with PRO assessments was >91% from BL until wk 27 in both arms. From BL to wk 27, LSM changes in GHS/QoL, PF, role functioning (RF), QLQ-HCC18 functional/domain scores and EQ5D-5L VAS were generally similar between the two arms (Table). Treatment with len + pembro delayed deterioration of patient-reported quality of life vs len + pbo (median TTD, 11.5 vs 4.3 mo; HR, 0.80; 95% CI, 0.65-0.98). LSM score difference (95% CI) was 0.5 (-2.5, 3.4) for GHS/QoL, -1.7 (-4.5, 1.1) for PF, -1.0 (-4.9, 2.9) for RF, 1.0 (-2.2, 4.3) for abdominal swelling, 2.0 (-1.2, 5.2) for fatigue, 2.0 (-1.0, 5.0) for pain, and 0.6 (-2.0, 3.2) for EQ5D-5L VAS. Median TTD in QLQ-C30 PF (5.7 vs 9.1 mo; HR, 1.14; 95% CI, 0.93-1.41) and QLQ-HCC18 abdominal swelling (NR vs NR; HR, 1.19; 95% CI, 0.89-1.60) /fatigue (2.9 vs 2.8 mo; HR, 0.98; 95% CI, 0.81-1.19)/pain (9.6 vs 7.9 mo; HR, 0.93; 95% CI, 0.75-1.16) was similar between the arms. Conclusions: Len + pembro generally preserved HRQoL vs len + pbo in 1L for aHCC. Combined with efficacy and safety results from LEAP-002, these findings support continued development of len + pembro in HCC. Clinical trial information: NCT03713593 . [Table: see text]

4095 Background: We conducted a multicenter prospective study (SURF trial) to compare the efficacy of surgery vs. radiofrequency ablation (RFA) for small hepatocellular carcinoma (HCC). SURF trial consisted of a randomized controlled trial (SURF-RCT) and a non-randomized prospective observational study (SURF-Cohort), including patients who did not consent to randomization. The initial report of the SURF-Cohort trial showed that recurrence-free survival (RFS) did not differ significantly between patients undergoing surgery and RFA. The focus of the present report was to assess the efficacy for overall survival (OS). Methods: The SURF-Cohort trial was a prospective multicenter study conducted in 49 institutions in Japan. Patients (aged between 29 and 79 years) with Child-Pugh scores ≤ 7, largest HCC diameter ≤ 3 cm, and ≤ 3 HCC nodules were considered eligible. Before the enrollment, both liver surgeons and hepatologists who perform RFA confirm that all the patients can be treated using both surgery and RFA. The co-primary endpoints were RFS and OS. OS was assessed at 5 years after the last accrual as per the protocol. Inverse probability of treatment weighted (IPTW) analysis was used to balance the characteristics of the groups. This trial is registered in UMIN000001796. Results: During 2009–2015, 782 patients were enrolled. After excluding ineligible patients, the surgery and RFA groups included 382 and 371 patients, respectively. In the surgery group, median platelet count (13.7x104 vs. 11.5x104, P< 0.01) was significantly greater, and the median largest HCC diameter was significantly greater (2.0 cm vs. 1.8 cm, P< 0.01) than in the RFA group. The median (range) follow-up period was 6.8 years in the surgery group and 6.7 years in the RFA group. The IPTW-adjusted OS did not differ significantly between the surgery and RFA groups: the 5-year OS, 79.7% vs. 79.3%; HR 0.98; 95% CI 0.75–1.30; P= 0.906. The analysis after long-term follow-up in the current report showed that RFS was not significantly different between the surgery and RFA groups: the 5-year RFS, 44.6% vs. 39.3%; HR 0.86; 95% CI 0.71–1.06; P= 0.155. Conclusions: SURF-Cohort trial revealed that OS and RFS after the IPTW adjustment were not significantly different between patients undergoing surgery and RFA for early stage HCC (≤ 3 cm and 3 nodules). Clinical trial information: UMIN000001796.

TPS627 Background: Pancreatic cancer is one of the most dismal cancers with few effective drugs. Both FOLFIRINOX (FFX) and gemcitabine plus nab-paclitaxel (GnP) showed superiority in overall survival (OS) to gemcitabine in phase III studies and became the standard of care for first-line chemotherapy. However, to date, there is no prospective randomized controlled studies comparing FFX and GnP. In addition, the original FFX has the problems of high toxicity and complicated administration, and we have been developing modified FFX with dose reduction of irinotecan and without bolus fluorouracil and S-IROX, which replaces continuous intravenous fluorouracil with orally administered fluoropyrimidine of S-1. The modified FFX showed median OS of 11.2months [95% confidence interval (CI), 9.0-not calculated] as good as the original FFX in a phase II study, and the S-IROX showed a high objective response rate (ORR) of 57.1% (95% CI, 34.0–78.2%) in a phase I study. This phase II/III study aims to confirm the superiority of modified FFX and S-IROX to GnP in metastatic or recurrent pancreatic cancer. Methods: The main eligibility criteria are metastatic or recurrent pancreatic cancer, histologically diagnosed as adenocarcinoma or adenosquamous carcinoma, no prior chemotherapy for pancreatic cancer, Eastern Cooperative Oncology Group Performance Status 0 or 1, and age 20-75 years old. Enrolled patients are randomized 1:1:1 to GnP, modified FFX, or S-IROX. GnP is consisted of nab-paclitaxel (125 mg/m2) and gemcitabine (1000 mg/m2) on days 1, 8, and 15, every 4 weeks, modified FFX is consisted of oxaliplatin (85 mg/m2), irinotecan (150 mg/m2), l-leucovorin (200 mg/m2), and fluorouracil (2400 mg/m2, 46-hour continuous infusion), every 2 weeks, and S-IROX is consisted of oxaliplatin (85 mg/m2), irinotecan (150 mg/m2) and S-1 (80 mg/m2, days 1-7), every 2 weeks. All regimens are administered until disease progression or unacceptable toxicity. The primary endpoint of the phase II part is the ORR to S-IROX with the null hypothesis of a threshold ORR of less than 20%, which decide whether to proceed to the phase III part in three arms or two arms (GnP and modified FFX). The primary endpoint of the phase III part is OS, and secondary endpoints are progression-free survival, ORR, incidence of adverse events (AEs) and serious AEs, and dose intensity. We calculated a sample of 732 patients to maintain 80% power at a one-sided alpha error of 2.5% in each comparison, and the hazard ratios of modified FFX and S-IROX versus GnP were estimated at 0.73 each. The study started patient accrual in April 2019 and 349 patients have been enrolled as of September 2021. Clinical trial registry: jRCTs031190009. Clinical trial information: jRCTs031190009.

4084 Background: In a phase 1b trial (NCT03006926), L + P had promising antitumor activity as first-line (1L) therapy in uHCC. We present exploratory biomarker analyses of circulating angiogenic factors and cytokines/chemokines related to the mechanism of action of L + P (ie, pharmacodynamic [PD] biomarkers), as well as biomarker correlations with clinical outcomes in patients (pts) with uHCC, from this trial. Methods: Pts received lenvatinib 12 mg/d (bodyweight [BW] >60 kg) or 8 mg/d (BW < 60 kg) PO + pembrolizumab 200 mg IV Q3W. Tumors were assessed using mRECIST or RECIST v1.1 per independent imaging review. Peripheral blood samples were collected before administration of study drug at baseline, cycle (C) 2, day (D) 1, C3D1, C4D1, and off-treatment. 43 Biomarkers were assayed in serum from 100 1L uHCC pts (excluding 4 pts from the dose-limiting toxicity part of the trial with prior sorafenib). Of these 43, 31 biomarkers (for which ≤20% of samples had measurements above/below the quantification limit of the assay) were included in the analyses. Changes in biomarker levels from baseline were evaluated via 1-sample Wilcoxon signed-rank test. Associations were explored between changes in biomarker levels and maximum tumor shrinkage (MTS) via the Spearman’s rank correlation test, objective response (OR; complete response + partial response) via the Wilcoxon rank sum test, and PFS via Cox regression analysis and log rank test. Data cutoff date for clinical endpoints was 7 August 2020. Results: Levels of PD biomarkers related to angiogenic signaling (VEGF increase/ANG2 decrease), FGF signaling (increase in FGF23/FGF19), and IFNγ signaling (increase in IFNγ, CXCL9/10/11) were changed significantly (adjusted P< 0.05) with L + P (C2D1–C4D1; except for FGF19 at C3D1). Significant decreases of TIMP1 and increases of MCP1 were observed at C4D1 during treatment; these were associated with greater MTS. Greater decreases in TIMP1 and greater increases in MCP1 were observed in pts with OR vs others. Changes in levels of the PD biomarkers ANG2, IL10, and VEGFR2 were found to be associated with PFS by dichotomized analysis. With tertile 2 cutoff, median PFS for pts in the group with greater decreases of ANG2 was 13.9 months vs 9.6 months for pts in the group with lesser decreases of ANG2 (unadjusted P= 0.002; HR 2.65, 95% CI 1.39–5.08). Conclusions: These are the first exploratory biomarker analyses for the single-arm study of L + P in pts with uHCC. Changes in serum biomarkers associated with angiogenic-, FGF-, and IFNγ-signaling pathways indicated target engagement of L + P. Decreases in TIMP1 and increases in MCP1 were associated with MTS and OR. Associations were found between longer PFS and a greater decrease in levels of ANG2. Angiogenesis inhibition and modulation of cancer immune response were observed with L + P. Further validation from independent studies is warranted. Clinical trial information: NCT03006926.

4093 Background: The initial report of the multicenter SURF trial (surgery vs. radiofrequency ablation [RFA] for small hepatocellular carcinoma [HCC]) showed that recurrence-free survival (RFS) did not differ significantly between patients undergoing surgery and RFA. The focus of the present report was to assess the effect on overall survival (OS). Methods: The SURF trial was a multicenter, open-label, randomized, controlled, phase 3 trial conducted in 49 institutions in Japan. Patients (aged between 29 and 79 years) with Child-Pugh scores ≤ 7, largest HCC diameter ≤ 3 cm, and ≤ 3 HCC nodules were considered eligible. Before enrollment, both liver surgeons and hepatologists who perform RFA confirm that all the patients can be treated using both surgery and RFA. Patients were then randomly assigned in a 1:1 ratio to undergo surgery or RFA, stratified by age, hepatitis-C virus infection, numbers of HCC, largest HCC diameter, and institution. The co-primary endpoints were RFS and OS. As per the protocol, RFS was reported previously at 3 years after the last accrual of patients. OS was planned at 5 years after the last accrual. This trial is registered in UMIN000001795. Results: During 2009–2015, 308 patients were enrolled. After excluding ineligible patients, the surgery and RFA groups included 150 and 152 patients, respectively. Baseline factors did not differ significantly between the groups. In both groups, 90% of patients had solitary HCC. The median largest HCC diameter was 1.8 cm (interquartile range, 1.5–2.2 cm) in the surgery group and 1.8 cm (interquartile range, 1.5–2.3 cm) in the RFA group. The median (range) follow-up period was 6.4 (0.4–10.8) years in the surgery group and 6.6 (0–10.7) years in the RFA group. OS did not differ significantly between the surgery and RFA groups as the 5-year OS (95% confidence interval [CI]) was 74.6% (66.5%–81.0%) in the surgery group and 70.4% (62.2%–77.3%) in the RFA group (hazard ratio (HR), 0.96; 95% CI, 0.64–1.43; P= 0.828). The analysis after long-term follow-up in the current report showed that RFS was not significantly different between the surgery and RFA groups: the 5-year RFS (95% CI), 54.7% (46.0%–62.5%) vs. 50.5% (42.1%–58.3%); HR 0.90; 95% CI 0.67–1.22; P= 0.498. Conclusions: SURF trial revealed that OS and RFS were not significantly different between patients undergoing surgery and RFA for small HCC (≤ 3 cm and 3 nodules). Clinical trial information: 000001795.

4017 Background: FOLFIRINOX, consisting of leucovorin (LV), fluorouracil (FU), irinotecan (IRI) and oxaliplatin (L-OHP), and GnP, consisting of gemcitabine (GEM) plus nab-paclitaxel (nPTX), have shown superior efficacy over GEM in patients (pts) with metastatic pancreatic cancer. Although several studies have reported the efficacy of FOLFIRINOX or GnP for pts with locally advanced pancreatic cancer (LAPC), no randomized controlled trial to compare the two regimens has been conducted in those pts. To select the most promising chemotherapy for LAPC, a randomized phase II selection design trial (JCOG1407) was conducted to compare between modified FOLFIRINOX (FOLFIRINOX with dose reduction of IRI and without bolus FU; Arm A) and GnP (Arm B) for pts with LAPC. Methods: In Arm A, 85 mg/m2 of L-OHP, 200 mg/m2 of l-LV, 150 mg/m2 of IRI, followed by 2,400 mg/m2 of continuous FU over 46 hours are infused every 2 weeks. In Arm B, 125 mg/m2 of nPTX followed by 1,000 mg/m2 of GEM are infused on days 1, 8, and 15 every 4 weeks. The primary endpoint was overall survival (the proportion of 1-year OS), and secondary endpoints included progression-free survival (PFS), distant metastasis-free survival (MFS) and response rate in pts with target lesions. The planned sample size was 124 pts to select more effective regimen in 1-year OS with a probability of at least 0.85 and to test the null hypothesis of 53% in 1-year OS with a one-sided alpha of 5% and 80% Results: From 2015 to 2019, a total of 126 pts was enrolled from 29 Japanese institutions, and were allocated to Arm A (n = 62) or Arm B (n = 64). The median (range) age was 66 (44-75) years and 58.7% were male. At the analysis, after a median (range) follow-up of 1.52 (0.55-3.99) years, 75 (59.5%) pts died. The proportion of 1-year OS was better in Arm B, 77.4% [95% CI 64.9–86.0] vs. 82.5% [95% CI 70.7–89.9], but 2-year OS was better in Arm A, 48.2% [95% CI 33.3–61.7] vs. 39.7% [95% CI 28.6–52.5]. Median OS was 2.0 years [95% CI 1.6-2.7] in Arm A and 1.8 years [95% CI 1.5-2.0] in Arm B. 1-year PFS for Arm A/B was 47.5 % [95% CI 34.5-59.4]/40.2% [95% CI 27.8-52.3], and 1-year MFS was 64.2 % [95% CI 50.9-74.8]/57.3% [95% CI 43.9-68.6]. Arm A was better OS in pts with CA19-9 1000 U/mL. Response rate was 30.9% [95% CI 19.1-44.8] in Arm A, and 41.4% [95% CI 28.6-55.1]) in Arm B. Incidences of grade 3-4 non-hematological toxicities for Arm A and Arm B were 66.1% and 67.2%, respectively. There was no treatment-related death. Conclusions: This study was the first randomized trial comparing the two regimens. The 1-year OS of the primary endpoint in GnP was better than mFOLFIRINOX, but mFOLFIRINOX achieved longer survival in 2-year OS. It is required to confirm longer OS and safety profiles which regimen should be selected as a standard regimen in LAPC. Clinical trial information: jRCTs031180085.

4519 Background: LEN is a multikinase inhibitor of VEGFR 1–3, FGFR 1–4, PDGFRα, RET, and KIT, approved for first line (1L) treatment of uHCC. PEMBRO, an anti-PD-1 monoclonal antibody, was granted accelerated approval for the treatment of patients (pts) with HCC after sorafenib therapy. We assessed the safety and efficacy of LEN + PEMBRO in uHCC. Methods: In this phase 1b trial (NCT03006926), pts received LEN 12 mg/day (bodyweight [BW] ≥60 kg) or 8 mg/day (BW

4581 Background: In parallel with a multicenter randomized controlled trial that reported an equal recurrence-free survival (RFS) of early-stage hepatocellular carcinoma (HCC) patients who underwent either surgery (SUR) or radiofrequency ablation (RFA), we also enrolled HCC patients who fulfilled the enrollment criteria but did not give consent to participate in the RCT. Methods: All patients gave informed consent to participate in this study. Inclusion criteria were as follows: primary HCC with less than or equal to 3 tumors, each measuring 3 cm or smaller; without vascular invasion or extrahepatic metastasis; Child-Pugh score of 7 or less; and ages between 20 and 79 years. The feasibility for both treatments was confirmed by a joint chart review by surgeons and hepatologists. The primary endpoint was RFS and overall survival. A pre-specified interim analysis was performed to compare RFS. Results: Between April 2009 and August 2015, 740 patients (371 in SUR, 369 in RFA) were enrolled from 49 participating hospitals in Japan. The SUR group had significantly fewer patients with chronic hepatitis C (56.6% vs. 69.4%), higher median value of platelet count (145 vs. 120 × 109/L), and more patients with > 2 cm tumors (49.9% vs. 27.9%); most patients had a single tumor (91.1% vs. 88.3%). During the median follow-up period of 5 years, tumor recurrence was observed in 192 of SUR and 218 of RFA with 3-year RFS being 66.0% and 61.7%, respectively ( P = 0.091). In subgroup analysis, RFS was significantly better in SUR in patients with ≤ 2 cm tumors (62.9% vs. 51.7% in 3 years; hazard ratio [HR] 0.72, 95% confidence interval [CI] 0.56-0.93; P = 0.014), whereas the difference was not significant in those with > 2 cm tumors (52.7% vs. 46.4%; HR 0.85, 95% CI 0.63-1.18; P = 0.34). The adjusted HR for RFS using inversed probability of treatment weighting was 0.89 (95% CI, 0.72-1.10; P = 0.287). Conclusions: The imbalance in patient characteristics reflected a real-world practice. Factors related to background liver disease rather than tumor characteristics might have a larger impact on the recurrence in early HCC. Clinical trial information: C000001796 .

TPS4152 Background: Len, an inhibitor of VEGF receptors 1-3, FGF receptors 1-4, PDGF receptor α, RET, and KIT, is approved for first-line treatment of unresectable HCC (uHCC) based on the open-label phase 3 REFLECT study in which len showed noninferior overall survival (OS) and significantly improved objective response rate (ORR), progression-free survival (PFS), and time-to-progression (TTP) vs sorafenib. In the phase 2 KEYNOTE-224 study of pembro (a PD-1 inhibitor) as second-line treatment of advanced HCC, pembro showed meaningful clinical efficacy in pts previously treated with sorafenib, with median PFS 4.9 mo, median OS 12.9 mo, and a manageable safety profile. In results from the phase 1b KEYNOTE-524 trial, len+pembro was well-tolerated, with promising antitumor activity in pts with uHCC. LEAP-002 is a phase 3 study to evaluate the safety and efficacy of len+pembro vs len+placebo as first-line therapy for advanced HCC. Methods: Eligible pts are ≥18 y and have HCC confirmed by radiology, histology, or cytology; ECOG PS 0/1; BCLC stage C or stage B disease not amenable to locoregional therapy or curative treatment approach; CP class A liver score within 7 days before study; and ≥1 measurable lesion by RECIST v1.1. Pts with past or ongoing HCV infection and those with controlled HBV are eligible. 750 pts will be randomized 1:1 to receive len 12 mg (body weight [BW] ≥60 kg) or 8 mg (BW 400 ng/mL; and ECOG PS 0/1. Primary end points are PFS per RECIST v1.1 by blinded independent central review (BICR) and OS. Secondary end points are ORR, duration of response, disease control rate, and TTP per RECIST v1.1 by BICR, efficacy per modified RECIST, pharmacokinetics, and safety. Imaging assessments will be performed Q9W on study. AEs will be graded per CTCAE v4.0 and monitored up to 90 days after last dose. Clinical trial information: NCT03713593.

4002 Background: Surgery (SUR) and radiofrequency ablation (RFA) are both known to be effective therapy for treating patients with small oligonodular hepatocellular carcinoma (HCC), however there is only insufficient evidence about which therapy is more preferred approach. This randomized controlled trial was designed to prospectively compare the efficacy of SUR and RFA as the first approach to primary HCC. Methods: In this open-label trial undertaken at 49 hospital in Japan, we recruited patients having primary HCC with tumor foci numbering less than 3, each measuring 3 cm or less, Child-Pugh score of 7 or less, ages between 20 and 79 year. Before randomization, technical and liver functional feasibility for both treatment arms were confirmed by joint chart review by surgeons and hepatologists. Patients were then randomly assigned in a 1:1 ratio to undergo SUR or RFA, stratified by age, infection of hepatitis-C virus, number of tumors, tumor size and institution. The primary endpoint was recurrence free survival (RFS) and overall survival (OS). Results: Between April 2009 and August 2015, total 308 patients were enrolled to this trial. Because of ineligibility 15 patients were excluded, therefore 145 patients underwent SUR and 148 patients underwent RFA finally. There was no perioperative mortality. Under the median follow-up of 5 years, the 3-year RFS of patients underwent SUR and RFA was 49.8%, 47.7%, respectively (hazard ration [HR] 0.96, 95% CI 0.72-1.28; p = 0.793). OS will be analyzed and published after two years. Conclusions: SUR and RFA were both safe therapeutic approaches and provided equally RFS for early stage HCC smaller than 3 cm. Clinical trial information: UMIN000001795.

407 Background: Several new combination therapies, including GEM plus nab-paclitaxel (GnP) and FOLFIRINOX (FFX) has been developed for treating pancreatic ductal adenocarcinoma (PDAC) in these ten years. We investigated trends in characteristics, treatment patterns, and outcomes of unselected patients with unresectable PDAC in real-life practice in Japan. Methods: We retrospectively reviewed the medical records of 1917 patients diagnosed as having unresectable or recurrent PDAC in multiple centers in our local area between January 2009 and April 2018. Results: The median age was 74, and 53.1% were men; 27.2% had locally advanced and 67.2% metastatic disease, and 5.6% had recurrences. Oncological therapy was administered to 1295 patients (67.6%): chemotherapy (n = 1161), chemo-radiotherapy (n = 117), or radiotherapy (n = 17); the remaining patients were treated with best supportive care. Of 100 patients diagnosed in 2009, 62.0% received GEM as first-line chemotherapy; whereas 56.8% of the 266 patients diagnosed in 2017 or 2018 received GnP, 20.3% GEM, and 15.4% FFX. The objective response rates of patients treated with GnP, FFX and GEM were 16.8%, 17.6%, and 5.1%, respectively. The median time to treatment failures were 3.9, 3.6, and 2.8 months, and the overall survivals were 11.5, 11.7, and 6.5 months after GnP, FFX, and GEM, respectively. Grade 3 or greater any hematological toxicity occurred in 53.7%, 64.7%, and 34.6% of the patients treated with GnP, FFX, and GEM, respectively. The treatment discontinuation rates due to adverse events were 17.2%, 14.9% and 22.9% in the patients treated with GnP, FFX and GEM, respectively. Conclusions: Chemotherapeutic protocols changed dramatically between 2009 and 2018 in Japan. Both GnP and FFX are well-tolerable and effective in real-life practice despite high frequent adverse events.

323 Background: Sorafenib, approved in Japan in 2009, is the first systemic therapy demonstrated to significantly improve overall survival (OS) in patients with advanced hepatocellular carcinoma (HCC). In Japan, hepatic arterial infusion chemotherapy (HAIC), which directly delivers high concentrations of cytotoxic agents to liver tumors, has been offered to patients with advanced HCC since before sorafenib was approved. HAIC is particularly used in patients without extrahepatic metastases (EHM). This study aimed to compare the outcomes of patients with advanced HCC who received HAIC and sorafenib. Methods: Consecutive patients with advanced HCC who received sorafenib or HAIC as the first-line systemic therapy were enrolled from 10 Japanese centers. The statistical analysis plan included pre-defined propensity score matching method and risk factors. All statistical analyses were performed by an independent biostatistician. Results: Between June 2009 and May 2016, 2006 patients were enrolled (sorafenib: 1465 patients, HAIC: 541 patients). The mean OS of patients with macrovascular invasion (MVI) and without EHM was significant longer in the HAIC group compared with the sorafenib group. After propensity score matching, there were 172 patients in each cohort. The OS was 9.1 months for the sorafenib group and 10.1 months for the HAIC group (hazard ratio [HR]: 0.668 [95% CI: 0.475–0.935], P = 0.018). There was no significant difference in OS between patients without both MVI and EHM. After propensity score matching, there were 76 patients in each cohort. The OS was 15.4 months for the sorafenib group and 12.2 months for the HAIC group (hazard ratio [HR]: 1.227 [95% CI: 0.699–2.155], P = 0.475). Conclusions: HAIC might be a potential initial treatment for patients with advanced HCC with MVI (without EHM). Currently, several new drugs appear clinically beneficial for patients with advanced HCC. Although this study only focused on sorafenib as the chemo-agent, additional studies should be conducted to confirm the benefits associated with HAIC in a limited population of patients with advanced HCC.

355 Background: The RESORCE trial proved survival benefit of regorafenib on advanced hepatocellular carcinoma (aHCC) in the limiting cohort. This study aimed to investigate the efficacy of hepatic arterial infusion chemotherapy (HAIC) as second-line treatment in patients with aHCC who are unfit for regorafenib. Methods: We studied consecutive 159 patients with aHCC treated with sorafenib in our institution from June 2009 to March 2018. We divided them into two groups according to the eligibility for RESORCE trial (Rego or non-Rego group) and investigated the efficacy of HAIC as post-sorafenib treatment. Moreover, we also analyzed the prognostic factors in each group. Results: When sorafenib treatment failed, 63 patients (39.6%) fulfilled key inclusion criteria of the RESORCE trial and were considered as fit for regorafenib treatment (Rego group). Among 96 remaining patients (non-Rego group), 34 patients (35.4%) were treated with HAIC after sorafenib. The objective response and the median progression-free survival of HAIC were 38.2% and 3.9 months, which were independent of the efficacy or treatment duration of sorafenib. The median survival of HAIC after sorafenib was 11.3 months which was significantly longer than that of 62 patients without HAIC (4.7 months; P = 0.023). Four factors, HAIC (HR 0.516), Child-Pugh A (HR 0.362), tumor size ≥35mm (HR 1.959), and AFP ≥200 (HR 1.728), were identified as independent prognostic factors after sorafenib in multivariate analysis. In Rego group, 14 patients were actually treated with regorafenib and their survival after sorafenib was longer than that of the patients without regorafenib and with HAIC (P < 0.01 and P < 0.01, respectively). In Rego group, regorafenib (HR 0.152), extrahepatic lesions (HR 2.160), and ≥10 intra-hepatic lesions (HR 2.099) were identified as prognostic factors after sorafenib in multivariate analysis. Conclusions: HAIC showed promising tumor response and patients’ outcome for the patients who are unfit for regorafenib. Further investigation is needed to compare HAIC with ramucirumab, cabozantinib or nivolumab.

4076Background: LEN is a multikinase inhibitor of VEGFR 1–3, FGFR 1–4, PDGFRα, RET, and KIT that showed noninferiority with respect to overall survival (OS) compared with sorafenib in the first-lin...

4067Background: The presence of portal vein tumor thrombosis (PVTT) in patients with hepatocellular carcinoma (HCC) is regarded as indicating an advanced stage, and liver resection (LR) is not reco...

229 Background: S-CUBE was a randomized, double-blind, phase III trial evaluating the efficacy and safety of S-1 in patients with sorafenib-refractory advanced hepatocellular carcinoma (HCC). The study’s primary outcome was presented at the 2015 ASCO Annual Meeting. Although S-1 did not significantly improve overall survival (OS) in all cohort (hazard ratio [HR] = 0.86; confidence interval [CI] = 0.67–1.10; P = 0.2201), we conducted predictive enrichment strategy analysis (PESA) to identify a patient population with a better response to S-1. Methods: Predictive enrichment strategy is a newly introduced concept proposed by the United States Food and Drug Administration, “to select a study population in which detection of a drug effect (if one is in fact present) is more likely than it would be in an unselected population. (Temple R. [2012] )” Therefore, in our study, PESA provides robust results and identifies advanced HCC patients who are more likely to respond to S-1. Clinically meaningful baseline characteristics were selected to create a scoring system; patients were ranked based on their scores, and the population with a better response was identified. Patient mapping was used to further characterize the population. Results: The full S-CUBE analysis set consisted of 333 patients, including 222 in the S-1 arm and 111 in the placebo. PESA and patient mapping identified 219 patients (65.8% of the total population) as the high-response patient population. High-response patients are classified as those with the following criteria; 1) TNM stage III, IVa, or IVb, 2) Child-Pugh class A, and 3) Levels of both the tumor markers are not high (AFP ≥ 400 ng/mL and PIVKA-II ≥ 10000 mAU/mL). In this population, the median OS of S-1 group was significantly longer than that of placebo group (426.0 days vs. 375.5 days; HR, 0.69; 95% CI, 0.51 to 0.93; P = 0.0156). Conclusions: PESA and patient mapping identified a high-response patient population to S-1. This statistically robust analysis demonstrated S-1 showed survival benefit for identified clinically-important population of sorafenib-refractory advanced HCC patients. Clinical trial information: JapicCTI-090920.

4075 Background: In the current AJCC/UICC staging system 7thed. for intrahepatic cholangiocarcinoma (ICC), tumor size was excluded and “periductal invasion” was added as a new factor determining T-...

4076 Background: Intra-arterial cisplatin therapy for advanced hepatocellular carcinoma (HCC) has been reported a favorable tumor-shrinking effect and long-term survivals even in pts with highly ad...

354 Background: Sorafenib has been used as the first-line treatment for advanced hepatocellular carcinoma (HCC), however, its efficacy and safety in Japanese patients (pts), especially those with Child-Pugh (CP) B cirrhosis, have not yet been fully examined. This study was conducted to evaluate the efficacy and safety of sorafenib in Japanese pts with HCC and CP B or CP A cirrhosis. Methods: The eligibility criteria were patients 1) with pathologically or clinically proven HCC, 2) with an ECOG performance status 0 to 2, 3) aged 20 to 79 years, 4) with measurable lesions, 5) with adequate hematological, renal and Child Pugh class A or B liver functions. Sorafenib was administered orally at the dose of 400 mg twice daily. Administration was continued until the detection of disease progression or appearance of unacceptable toxicity. The primary endpoint of the study was progression-free survival (PFS), and the secondary endpoints included objective response, overall survival (OS), and toxicity. Results: Forty CP A pts and 12 CP B pts were enrolled between April 2010 and January 2012. The median PFS in the CP A pts was 3.3 months (M) and that in the CP B pts was 3.2 M. Among the pts with CP A, there was one patient with confirmed complete response (2.5%), 3 pts with partial response (7.5%), and 19 pts (47.5%) with stable disease (SD). Among the pts with CP B, there were no treatment responses, and 8 (66.7%) pts had SD. The median overall survival in the CP A pts was 13.4 M and that in the CP B pts was 7.4 M. With regard to toxicities, fewer CP A pts experienced grade 3/4 toxicities than CP B pts (77.5% vs. 91.6%). The grade 3/4 toxicities in the CP A and B pts, respectively, included thrombocytopenia (10% and 25%), hand foot skin reaction (27.5% and 16.7%), Erythema multiforme (0% and 16.7%), and upper gastrointestinal bleeding (0% and 16.7%). There were no treatment-related deaths in either group of patients. Conclusions: This study shows that sorafenib is effective and well-tolerated in Japanese patients with HCC and Child Pugh class A liver cirrhosis, consistent with previous reports. The outcome was poorer and severe toxicities were more frequent in patients with Child Pugh B cirrhosis than in those with Child Pugh A cirrhosis. Clinical trial information: 000002972.

165 Background: HCC has a high recurrence rate and poor prognosis; however, no effective treatment has been established for preventing recurrence after resection or ablation. Peretinoin, an acyclic retinoid, of 600 mg/day reduced HCC recurrence and a strong effect after two years suggested that it mainly suppressed de novo carcinogenesis (ASCO2010). Liver impairment causes non- HCC-related deaths or adverse events, and can potentially mask drug efficacy. The size of the main HCC nodule associates vascular invasion and intrametastasis. To evaluate peretinoin efficacy and assess its mechanism of action, we performed subgroup analysis in CP-A patients who had undergone curative therapy for a small tumor. Methods: Eligibility criteria included HCV positive, CP-A or B, and prior curative resection or ablation for conventional HCC. The patients were randomized into groups to receive peretinoin 600 mg/day (high dose, H), peretinoin 300 mg/day (low dose, L), or placebo (P). The patients were grouped according to the CP grade and the size of the major tumor Results: Of the 401 patients, 310 showed CP-A, and 144 showed CP-A and tumor size Conclusions: Subgroup analysis with CP-A reinforced the peretinoin efficacy in preventing HCC recurrence. In addition, the CP-A and tumor size [Table: see text]

4030 Background: HCC is well known for hypervascular tumors and currently clinical trials for many antiangiogenic drugs are conducted. TSU-68 is an oral compound which inhibits VEGFR, PDGFR and FGFR, and it showed promising efficacy with a high safety profile, even in heavily pretreated HCC pts with Child-Pugh B liver cirrhosis (Kanai et al., 4589, ASCO 2008). TACE is an internationally established standard treatment for pts with intermediate stage HCC. Most pts will have intrahepatic recurrence after TACE, and VEGF level has been reported to be associated with recurrence. We conducted a randomized phase II trial to assess the efficacy and safety of the TSU-68 after TACE. Methods: This study was an open-label, multicenter, randomized phase II trial. Pts who were not indicated to surgical treatment or local treatment such as radiofrequency ablation, absence of portal vein thrombosis and extrahepatic metastases, and had either Child-Pugh A or B liver function were randomly assigned to group A with TSU-68 40...

4588 Background: This randomized phase II trial compared the response rate (RR) of interferon (IFN) combined 5-fluorouracil (5FU) plus cisplatin (CDDP) hepatic arterial infusion (HAI) with IFN combined 5FU HAI in advanced hepatocellular carcinoma. Methods: Patients (pts) with measurable histologically or radiologically confirmed advanced hepatocellular carcinoma (major vascular invasion and/or bilobular multiple ≥5 nodules) were randomized into 2 groups. Arm A (n=57): continuous 5FU HAI (300mg/m2 day1–5, day8–12), CDDP HAI (20mg/m2, day 1, day 8 for 1.5 hours). IFN alpha-2b (3M IU/body) was administrated intramuscularly 3 times per week for 4 weeks. The treatment cycles repeated for 6 weeks. Arm B (n=57): IFN combined 5FU HAI with same dose without CDDP infusion. Treatment was continued until disease progression. The primary endpoint was RR. The secondary endpoints were overall survival (OS), time to progression (TTP) and toxicity. Results: Results for responses are presented for 109 pts and toxicity for 114 pts. The best overall response rate (RR) was 46% for group A and 25% for group B. This included 1 (2% of group A) vs. 3 (6% of group B) complete responses, and 25 (44% of group A) vs. 11 (19% of group B) partial responses. Fifteen pts in group A (26%) vs. 19 pts (33%) in group B had stable disease and 13 pts (23%) vs. 22 (39%) in respectively group A and B progressed while on treatment. RR was significantly higher in group A (p=0.02). The median TTP was 6.5 ± 2.0 months (mo) for group A vs. 3.3 ± 2.0 mo for group B (p=0.005). The median OS was 17.6 ± 3.2 mo for group A vs. 10.5 ±2.3 mo for group B (p=0.38). The median OS was 13.7 ± 4.9 mo for both groups. Grade 3/4 toxicity occurred in 65.8% of pts. Hematological toxicity was common and occurred in 53.5%. The thrombocytopenia and infusion port-related toxicity occurred more frequently in group A. Conclusions: IFN combined 5FU plus CDDP HAI shows higher antitumor activity and longer TTP than IFN combined 5FU HAI. Although OS is also longer without significant difference, these results show the clinical efficacy of additional CDDP to IFN combined 5FU HAI. No significant financial relationships to disclose.