Search

PURPOSE This phase I/II study evaluated tremelimumab (anticytotoxic T-lymphocyte–associated antigen-4 monoclonal antibody) and durvalumab (antiprogrammed death ligand-1 monoclonal antibody) as monotherapies and in combination for patients with unresectable hepatocellular carcinoma (HCC), including a novel regimen featuring a single, priming dose of tremelimumab (ClinicalTrials.gov identifier: NCT02519348 ). PATIENTS AND METHODS Patients with HCC who had progressed on, were intolerant to, or refused sorafenib were randomly assigned to receive T300 + D (tremelimumab 300 mg plus durvalumab 1,500 mg [one dose each during the first cycle] followed by durvalumab 1,500 mg once every 4 weeks), durvalumab monotherapy (1,500 mg once every 4 weeks), tremelimumab monotherapy (750 mg once every 4 weeks [seven doses] and then once every 12 weeks), or T75 + D (tremelimumab 75 mg once every 4 weeks plus durvalumab 1,500 mg once every 4 weeks [four doses] followed by durvalumab 1,500 mg once every 4 weeks). Safety was the primary end point. Secondary end points included objective response rate (ORR) by Response Evaluation Criteria in Solid Tumors v1.1 and overall survival; exploratory end points included circulating lymphocyte profiles. RESULTS A total of 332 patients were enrolled (T300 + D, n = 75; durvalumab, n = 104; tremelimumab, n = 69; and T75 + D, n = 84). Tolerability was acceptable across arms, with grade ≥ 3 treatment-related adverse events occurring in 37.8%, 20.8%, 43.5%, and 24.4%, respectively. Confirmed ORRs (95% CI) were 24.0% (14.9 to 35.3), 10.6% (5.4 to 18.1), 7.2% (2.4 to 16.1), and 9.5% (4.2 to 17.9), respectively. An early expansion of CD8+ lymphocytes was associated with response across arms, with highest proliferating CD8+ lymphocyte levels occurring in the T300 + D arm. The median (95% CI) overall survival was 18.7 (10.8 to 27.3), 13.6 (8.7 to 17.6), 15.1 (11.3 to 20.5), and 11.3 (8.4 to 15.0) months in the T300 + D, durvalumab, tremelimumab, and T75 + D arms, respectively. CONCLUSION All regimens were found to be tolerable and clinically active; however, the T300 + D regimen demonstrated the most encouraging benefit-risk profile. The unique pharmacodynamic activity and association with ORR of the T300 + D regimen further support its continued evaluation in HCC.

455 Background: Malignant ascites occurs frequently in patients with gastric cancer (GC) and colorectal cancer (CRC) with peritoneal metastasis (PM). The presence of PM and malignant ascites have been independently reported to confer resistance to systemic therapy and poorer prognosis. However, the occurrence of malignant ascites as a function of increasing peritoneal carcinomatosis burden has been less studied and reported. Methods: We reviewed prospective cohorts of gastric cancer and colorectal cancer patients with PM. The first cohort was a prospective group of GCPM patients receiving bi-directional systemic and peritoneal-directed therapies (catheter-based intraperitoneal chemotherapy and/or pressurized intraperitoneal aerosol chemotherapy respectively) at a tertiary oncology center in Singapore. Clinico-pathological data, including Peritoneal Cancer Index (PCI) and the presence or absence of ascites based on diagnostic laparoscopy, was collected. To orthogonally validate the hypothesis of malignant ascites as a function of peritoneal carcinomatosis burden in gastrointestinal malignancies, we studied the relationship between ascites, PM and survival in an independent cohort of metastatic CRC patients eligible for first-line systemic treatment enrolled in two randomized clinical trials in different Italian cancer centers. Results: In the first cohort of 82 patients with GCPM, the median PCI was 19 (inter-quartile range (IQR) 8 – 26) in the group with ascites and 3 (IQR 1 – 12) in the group without ascites (p=0.005), suggesting ascites occurs with higher PM burden. The median overall survival (OS) was poorer in patients with ascites (13.4 vs 16.4 months, HR 1.6, 95% CI 0.9 – 2.8, p=0.082). The median OS in patients with PCI of 7. In the validation cohort of 900 CRC patients, presence of malignant ascites and PM resulted in poorer survival compared to patients with PM and no malignant ascites (HR for OS PM with ascites vs PM without ascites 2.01 (95% CI 1.37 - 2.96), p

3584 Background: Trifluridine/tipiracil (TAS-102) is currently approved as third-line treatment in metastatic colorectal cancer (mCRC). However, there is paucity of real-world data on the tolerability and efficacy with biweekly dosing as monotherapy or in combination with bevacizumab. In this study, we present our center’s experience with biweekly TAS-102 with or without bevacizumab in mCRC patients (pts). Methods: We performed a single center retrospective observational study of pts receiving TAS-102 between 2018 and 2021. Results: A total of 83 pts were included (53 men, 30 women), with a median age of 64 years. Majority of pts were treated with TAS-102 in the 3rd-line (48.2%) and 4th-line (28.9%) setting. Almost all (94.0%) were of ECOG ≤ 1 at the initiation of treatment. The mean number of cycles administered was 3.8 and bevacizumab (5mg/kg on Day 1, every 2 weeks) was used in combination with TAS-102 in 18 pts (21.7%). Majority of pts (84.3%) were given TAS-102 using the biweekly regimen (35mg/m2 BD, on Day 1-5 and 15-19, q28 days) rather than standard regimen (35mg/m2 BD, on Day 1-5 and 8-12, q28 days) following a change in institutional practice. Fifteen pts (18.1%) had their initial dose reduced to 30mg/m2 BD at prescriber's discretion. Median PFS and OS were 2.37 and 10.15 months, respectively. In terms of tolerability, fatigue (any grade, 42.2%) and neutropenia (any grade, 44.6%) were the two most common adverse events reported. Grade 3 or higher neutropenia and febrile neutropenia were 16.9% and 3.6%, respectively. Dose reduction during treatment was required in 15 pts (18.1%), dose delay in 31 pts (37.3%) and six pts (7.2%) discontinued treatment due to toxicity. More than half (54.2%) had at least an additional line of therapy (regorafenib, clinical trials or re-challenge previous chemotherapy agents) following disease progression with TAS-102. Conclusions: We report the largest real-world experience with biweekly TAS-102. Our pts treated with TAS-102 had comparable median PFS to the RECOURSE cohort but with significantly better OS as more than half continued to receive treatment. Biweekly dosing of TAS-102 with or without bevacizumab is well tolerated with significantly lower rates of Grade 3 neutropenia compared to the published data in the literature.

570 Background: Gemcitabine-based chemotherapy is an approved therapy for treatment-naïve metastatic pancreatic ductal adenocarcinoma (PDAC). Inherent resistance to gemcitabine inevitably leads to cancer progression and shorter survival. It has been demonstrated that bacteria are a component of the PDAC tumor microenvironment and may play a critical role in mediating resistance to chemotherapy. Hence, the coadministration of an antibiotic to chemotherapy may potentiate antitumour drug responses. We conducted a pilot study to assess the efficacy and safety of ciprofloxacin plus gemcitabine-based chemotherapy in patients with treatment naïve metastatic PDAC. We also aimed to study gut microbiome changes during treatment with ciprofloxacin. Methods: This was a single arm study conducted at the National University Cancer Institute, Singapore. Patients (pts) with histologically confirmed metastatic PDAC were treated with nab-paclitaxel (125 mg/m2) and gemcitabine (1000 mg/m2) on days 1, 8, and 15 every 4 weeks, in combination with oral ciprofloxacin 500 mg twice daily. Treatment was continued until disease progression or intolerable toxicity. DNA extraction, 16S rRNA amplification and sequencing for bacteria were performed on pre- and post-treatment stool samples. Primary endpoint of this study was response rate and safety of the treatment combination. Secondary endpoints included progression free survival (PFS), overall survival (OS) and microbiome changes after treatment with ciprofloxacin. Results: From Mar 2019 – Feb 2021, 8 pts were recruited. Median age was 71 years old. Best response was stable disease in 5 (62.5%) pts and 3 pts had progressive disease (PD). Median PFS and OS were 4.3 and 15.4 months, respectively. 2 pts developed grade 4 neutropenia and 1 pt had grade 3 febrile neutropenia. Rash was common with 50% of pts developed grade 1/2 rash. No additional toxicities from ciprofloxacin were observed. Preliminary stool analysis showed significant differences in individual bacterial strains across all timepoints in the PD vs. SD groups. There was also an increased abundance in gammaproteobacteria resistant to ciprofloxacin treatment in pts with PD. Conclusions: Gemcitabine-based chemotherapy plus ciprofloxacin demonstrated clinical activity and acceptable safety profile in PDAC. Our study also highlighted that gut microbiome may play a critical role in mediating resistance to chemotherapy. Clinical trial information: NCT04523987.

439 Background: Standard first-line treatment for advanced biliary tract cancer (BTC) with gemcitabine plus cisplatin confers only modest benefits (objective response rate [ORR], 26%; Valle et al. N Engl J Med 2010;362:1273-1281) and patient prognosis is poor. Therefore, new first-line strategies are urgently required. Human epidermal growth factor receptor (HER) tyrosine kinases are commonly overexpressed in BTC. Varlitinib is a reversible small molecule pan-HER inhibitor. Methods: This was a multicentre, phase 1B/2 study in patients with advanced BTC who had not received prior systemic therapy. Using a modified 3+3+3 escalation design in phase 1B, patients received oral varlitinib twice daily (BID) every day (starting dose 200 mg BID) plus gemcitabine/cisplatin on Days 1 and 8 in a 3-week cycle. The primary objective was to determine the maximum tolerated dose (MTD) and safety profile of varlitinib in combination with gemcitabine plus cisplatin. Patients were evaluable for MTD if they had varlitinib treatment compliance of ≥80% and received 2 doses of chemotherapy in Cycle 1, or if they experienced a dose-limiting toxicity (DLT) in Cycle 1. Preliminary efficacy was a secondary objective. Results: Overall, 23 patients were enrolled in Korea (3 sites), Singapore (1 site), and Taiwan (1 site). Patient characteristics at baseline: male, 52%; median (range) age, 66 (47-82) years; ECOG PS, 0 (43%)/ 1 (57%); primary tumor site, intrahepatic bile duct (43%)/ extrahepatic bile duct (22%)/ gallbladder (22%)/ ampulla of Vater (13%). In total, 11 and 12 patients were included in the varlitinib 200 mg and 300 mg cohorts, of whom 9 and 7 were evaluable for MTD, respectively. DLTs were observed in 2 patients in the 200 mg cohort. Prior to study termination, the 300 mg dose was being assessed with 1 patient experiencing a DLT. In the overall population (n = 23), 36% of patients in the 200 mg cohort and 67% in the 300 mg cohort had at least one grade ≥3 treatment-related adverse event. Grade ≥3 treatment-related investigations abnormalities and blood and lymphatic system disorders were reported in 27% and 18% of patients in the 200 mg cohort, and 25% and 25% of those in the 300 mg cohort, respectively. Among all 23 patients, 8 had a partial response and 12 had stable disease for ≥12 weeks, giving an ORR of 35% and a disease control rate of 87%. The median (Q1-Q3) duration of objective response was 4.0 (2.8-11.0) months. At the time of analysis, 17 patients had progression events and 6 were censored. The median (Q1-Q3) progression-free survival was 6.8 (5.1-10.7) months. Conclusions: Varlitinib combined with gemcitabine plus cisplatin was well tolerated and associated with preliminary anti-tumor activity in patients with treatment-naïve advanced BTC. Clinical trial information: NCT02992340.

4090 Background: Evidence suggests that hyperactivated fibroblast growth factor 4 (FGFR4) signaling pathway leads to enhanced tumor growth. Targeting FGFR4 may have therapeutic benefit in tumors with altered FGF19 signaling. A phase I study (NCT02834780) was undertaken to assess H3B-6527, a highly selective covalent FGFR4 inhibitor, in patients with HCC/ICC. Methods: Adults with advanced HCC/ICC, ECOG PS 0-1, well compensated liver function, who progressed after > one prior therapy, received H3B-6527 po daily (QD) or twice-daily (bid) on a 21-day cycle following a 3+3 design. Doses ranged from 300-2000mg QD or 500-700mg BID. Patients in dose escalation were treated regardless of FGF19 status. Patients in expansion had FGF19+ tumors by mRNA testing. Adverse events (AEs), and pharmacokinetics (PK) were assessed. Response was determined by RECIST 1.1/mRECIST imaging every 6 weeks. Results: Study enrollment is complete at 128 patients. Ninety HCC patients were treated (QD = 48, bid = 42). ICC enrollment was suspended after 38 patients due to limited efficacy. No dose-limiting toxicities were seen and no grade 4-5 treatment related AEs have been observed. Recommended Phase II dose for H3B-6527 is 1000mg QD based upon safety, efficacy, and PK data. Grade 3 TEAEs have occurred in 12.5% of patients on QD dosing. Treatment related TEAEs were seen in 62.5% of patients on the QD schedule, with diarrhea (45.8%), fatigue (12.5%), and nausea (12.5%) most frequent. Drug discontinuation due to AEs for QD dosing was 8.3%. Interim data analysis shows that, for HCC patients with >2 prior lines of therapy treated on QD schedule, overall survival was 10.6m, progression-free survival 4.1m, overall response rate 16.7% (all partial responses), and clinical benefit rate 45.8% (responders + durable stable disease >17 weeks). H3B-6527 Cmax and AUC were lower at 300 mg dose but then similar across 500–2000 mg doses. Following oral administration of 1000 mg fasted, H3B-6527 plasma concentration reached peak at a Tmax of ̃2-3 hours and then decayed exponentially, with terminal half-life of ̃4-5 hours. There was no accumulation following QD dose. Dosing with food did not meaningfully change H3B-6527 plasma exposure. Conclusions: H3B-6527 was well tolerated and demonstrated a favorable toxicity and safety profile and encouraging clinical activity in heavily pretreated HCC patients. Final trial results will be presented at conference. Clinical trial information: NCT02834780.

265 Background: Treatment options for cholangiocarcinoma (CCA) after progression on first-line gemcitabine-based therapy are limited. Fibroblast growth factor receptor 2 ( FGFR2) gene fusions occur in 13–17% of intrahepatic CCA. A single-arm, phase II study (NCT02150967) evaluated infigratinib, an ATP-competitive FGFR1–3-selective oral tyrosine kinase inhibitor, in previously-treated advanced CCA with FGFR fusions/rearrangements. Methods: Adult patients with advanced/metastatic CCA with progression on ≥1 line of systemic therapy received infigratinib 125 mg orally for 21 days of each 28-day cycle until unacceptable toxicity or disease progression. All patients received prophylaxis with the oral phosphate binder sevelamer. Primary endpoint: objective response rate (ORR) by independent central review per RECIST v1.1, with duration of response (DOR). Secondary endpoints: progression-free survival (PFS), disease control rate, overall survival, safety, pharmacokinetics. Approximately 160 patients are planned (120/20/20 patients in Cohorts 1/2/3). This analysis focuses on Cohort 1 (patients with FGFR2 gene fusions or rearrangements without receiving a prior FGFR inhibitor). Results: As of 31 March 2020, 108 patients, including 83 (77%) with FGFR2 fusions, received infigratinib: median age 53 years (range 23–81 years); 54% had received ≥2 prior treatment lines. Median follow-up was 10.6 months (range 1.1–55.9 months). 96 patients (88.9%) discontinued treatment (12 ongoing). Centrally reviewed ORR was 23.1% (95% CI 15.6–32.2) including 1 CR and 24 PRs; median DOR was 5.0 months (range 0.9–19.1 months). Among responders, 8 (32.0%) patients had a DOR of ≥6 months. Median PFS was 7.3 months (95% CI 5.6–7.6 months). Prespecified subgroup analysis: ORR was 34% (17/50) in the second-line setting and 13.8% (8/58) in the third-/later-line setting (3–8 prior treatments). Most common treatment-emergent adverse events (TEAEs, any grade) were hyperphosphatemia (76.9%), eye disorders (67.6%, excluding central serous retinopathy/retinal pigment epithelium detachment [CSR/RPED]), stomatitis (54.6%), and fatigue (39.8%). CSR/RPED occurred in 16.7% of patients (including 1 G3 event; 0 G4). Other common grade 3/4 TEAEs were stomatitis (14.8%; all G3), hyponatremia (13.0%; all G3), and hypophosphatemia (13.0%; 13 G3, 1 G4). Conclusions: Infigratinib is associated with promising anticancer activity and a manageable AE profile in patients with advanced, refractory CCA with an FGFR2 gene fusion or rearrangement. A phase III study of infigratinib versus gemcitabine/cisplatin is ongoing in the front-line setting (NCT03773302). Clinical trial information: NCT02150967.

TPS356 Background: The FGFR family plays an important role in cholangiocarcinoma, with FGFR2 gene fusions detected in about 15% of patients with cholangiocarcinoma. Infigratinib is an FGFR1–3-selective oral tyrosine kinase inhibitor under evaluation in multiple indications including front-line and pre-treated cholangiocarcinoma. CBGJ398X2204 is an ongoing phase II study evaluating the efficacy of single-agent infigratinib in patients with advanced or metastatic cholangiocarcinoma with FGFR genetic alterations who have received prior gemcitabine. Methods: Study CBGJ398X2204 consists of 3 cohorts and patients in all cohorts receive oral infigratinib once daily for 21 days of a 28-day treatment cycle. Treatment will continue until progressive disease, intolerance, withdrawal of consent, or death. Cohort 1 includes patients with FGFR2 gene fusions or translocations. Cohort 2 includes patients with FGFR genetic alterations other than FGFR2 gene fusions (patients in both Cohorts 1 and 2 must not have received any prior FGFR inhibitors). Cohort 3 includes patients with FGFR2 gene fusions who have received prior treatment with a selective FGFR inhibitor other than infigratinib. The primary endpoint is objective response rate (ORR, RECIST v1.1 per central review). Secondary endpoints include overall survival and overall response rate (per investigator). Safety, pharmacokinetics, and exploratory genetic alterations/biomarkers will also be measured. The study was initiated in 2014 and has a planned enrollment of up to 160 patients across all 3 cohorts (120 in Cohort 1, 20 in Cohort 2, and 20 in Cohort 3). Cohort 1 has completed enrollment and findings from this Cohort are the focus of a separate abstract submitted to the meeting. Results are not currently available from Cohorts 2 and 3 (trial in progress). Clinical trial information: NCT02150967.

165 Background: The addition of intraperitoneal (IP) paclitaxel (PTX) to systemic chemotherapy comprising taxane/fluoropyrimidine doublet has shown promising results for patients with gastric cancer (GC) and peritoneal metastases (PM). However, this has not been studied in combination with platinum/fluoropyrimidine doublet which is the current standard-of-care for metastatic GC. We conducted a prospective phase 2 trial of IP PTX with capecitabine and oxaliplatin (XELOX) in patients with GCPM. Methods: The trial enrolled 44 patients with GCPM who received treatment comprising IP PTX (40mg/m2 on day 1,8), PO capecitabine (1000mg/m2 twice daily from day 1-14) and IV oxaliplatin (100mg/m2 on day 1) in 21-day cycles. Patients with synchronous GCPM were eligible for conversion surgery comprising radical gastrectomy if they had good response after chemotherapy, negative cytology on 2 consecutive peritoneal fluid assessments, no extraperitoneal metastasis and no peritoneal disease during surgery. The primary endpoint was overall survival and secondary endpoints were progression-free survival and safety. Outcomes from the trial were also compared with a retrospective cohort of 39 patients with GCPM who received identical systemic chemotherapy (SC) comprising platinum/fluoropyrimidine agents alone. Results: The median OS for the IP and SC groups was 14.6 and 10.6 months (HR 0.44; 95% CI, 0.26-0.74; P=0.002). The 1-year OS was 67.8% in the IP group and 32.3% in the SC group (Logrank p

e24095 Background: The chemo-brain effect associated with systemic chemotherapy results in cognitive disturbances impacting the capacity to engage in tasks and quality of life (QOL). Patients with colorectal cancer (CRC) who receive adjuvant chemotherapy generally have long survival times. The long-term effect of chemotherapy on cognition is uncertain. We aimed to ascertain the frequency of long-term cognitive impairment using neuropsychological assessments and correlating with neuroimaging. Methods: In this prospective pilot study, patients (n=22) with stage II to III CRC between 45 to 75 years old, who were planned to receive adjuvant chemotherapy, were recruited. 3 visits were scheduled for each subject – pre-chemotherapy (V1), at 1 month post chemotherapy (V2), and at 6 months post chemotherapy (V3). Serial tests were performed – the Cambridge Neuropsychological Test Automated Battery (CANTAB), QOL questionnaires (Hospital Anxiety and Depression Scale (HADS), Perceived Deficits Questionnaire (PDQ), EORTC QLQ-C30, FACT-ES), 3 item pocket smell test, functional PET/MRI brain imaging, and blood biomarker studies. Results: 18/22 subjects (13 male, 5 female) had completed tests at all 3 visits; the median age was 62 years (range 51 – 69). 9/18 had an initial decline (median -0.033) of Rapid Visual Information processing (RVP) at V2; 3/9 showed improvement to baseline at V3. 8/18 had a persistent decline in RVP scores at V3 (median -0.054). This was associated with increased HADS depression scores (mean 3.63 at V2 vs 4.63 at V1), worsening attention scores (mean 4.38 at V3 and 3.63 at V1), prospective memory scores (mean 3.75 at V3 vs 3.38 at V1), and total scores (mean 14.63 at V3 vs 13.75 at V1) on the PDQ. 7/18 had an increase in Paired Associates Learning (PAL) errors (median +6) at V2. 3/7 improved to baseline at V3, while 4/7 continued to have a persistent decline. PAL scores were not associated with worsening retrospective or prospective PDQ memory scores, changes in HADS depression or EORTC QLQ-C30 scores. There was no difference in baseline CANTAB scores for patients reporting declining vs stable QLQ-C30 scores. Conclusions: Only half of patients with initial RVP A and PAL decline improved at 6 months post chemotherapy. Further efforts should be placed to identify those at risk of poor recovery, and develop strategies to manage the chemo-brain effect. The correlation of cognitive decline with neuroimaging will be presented in the final analysis.

1019 Background: Preclinical studies show cross talk between RET and estrogen receptor, with at least additive treatment (Tx) effect of Len, a RET inhibitor, with Let. Our previous work concluded a recommended phase 2 dose (RP2D) of Len 14mg daily and Let 2.5mg daily (Lim, ASCO 2019). We present efficacy data from dose escalation and expansion cohorts. Methods: Safety, tolerability and efficacy data of MBC patients in both dose escalation (Len dose level 1 [DL]:20mg, DL -1:16mg and DL -2:14mg) and expansion (Len 14mg) cohort of this phase Ib/II study of combination Len+Let study was analysed. Patients were treated with single-agent Len for 2 weeks, followed by Len+Let until disease progression (PD). Serial tumor biopsies at baseline, after Len alone, 4 weeks post Len+Let, and upon PD, were sequenced for 440 genes with the ACTOnco+ platform. Results: A total of 33 pts (DL1 6pts, DL-1 6pts, DL-2 + expansion 21pts) with median 5 lines of prior Tx (range 0-11) were enrolled; 87.9%, 75.8%, and 75.8% had prior endocrine therapy (ET), ET+CDK4/6 inhibitor (i), and chemotherapy (CT) respectively. Objective response rate (ORR), disease control rate (DCR) ≥6 months (m), median duration of response (DOR), and percentage progression-free (PPF) at 12m were 33.3%, 45.5%, 11.5m (range 6.3-22.4), and 27.2% respectively. Among patients who previously progressed on CDK4/6i (n = 25), ORR, DCR ≥6m, median DOR, and PPF at 12m were 24.0%, 40.0%, 13.7m (range 6.3-18.2), and 12.0% respectively. Of note, 3/25 (12%) patients had durable response to Len+Let lasting ≥12m, despite having only modest PFS on ET+CDK4/6i (3, 7, and 12 months respectively). Most frequent all-grade toxicities (tox) were HTN (n = 15, G3:15), hypothyroidism (n = 20, G3:0) and fatigue (n = 13, G3:2), with no G4/5 tox. No new toxicity signals were observed compared to dose escalation phase. Pre-treatment tumor molecular profiling showed responders to be more likely to harbor NEFH, USH2A and PTCH1 mutations, while non-responders were more likely to carry PIK3R1, APC and PALB2 mutations. Sequencing of serial biopsies showed downregulation of BRD4, PTCH1, KIT, NTRK1 and CREBBP after Len treatment. Conclusions: Len+Let showed significant anti-tumor activity with meaningful duration of response, even in pts who failed prior CT or ET+CDK4/6i. The results support further investigation in randomized studies. Tumor profiling identified mutations associated with response and insights on molecular effects of lenvatinib. Clinical trial information: NCT02562118 .

3561 Background: Somatic profiling of MBC has highlighted actionable mutations and driven trials of matched targeted therapy (tx). Previous phase I studies have reported improved outcomes following matched therapies with tumour molecular profiles. Here, we review next generation sequencing (NGS) and treatment outcomes of Asian MBC patients (pts) in the phase I unit of a tertiary centre. Methods: Pts with MBC referred to a phase I unit underwent NGS (n = 152). Tumour tissue was sequenced via the amplicon based Ion Ampliseq Cancer (IAC) v2 (50 genes) platform from 2014-2017 prior to institutional change to Foundation Medicine 1 (FM1) (324 genes) 2017-2019. Patients were counselled on findings and enrolled onto matched therapeutic trials where available. Results: NGS was successfully performed in 107 pts (IAC 46%, FM1 54%) of which tumour subtypes include hormone receptor positive 63%, triple negative breast cancer (TNBC) 28% and Her2 positive 19%. Median lines of prior tx for MBC was 4 (range 0-12). 89% had prior chemotherapy (CT), 57% prior endocrine therapy (ET). 72/107 (67%) sequenced patients had further treatment and 18 (25%) were matched to tx based on NGS findings (15 clinical trial, 3 off trial). Matching rates on both NGS platforms were similar (IAC 22% vs FM1 28%). Mutated pathways with potential matched tx included PIK3CA/AKT/PTEN (52%), DNA damage response (DRR) (15%), and FGFR (11%) pathways. PIK3 mutations were seen in 43% and associated with higher number of metastatic sites (p = 0.03); most prevalent aberrations were PIK3CA H1047R (41%) and PIK3CA E542K (13%). Matched cases were more heavily pretreated (mean lines of prior tx 5.3 matched vs 3.7, unmatched p = 0.05), and showed a median progression free survival (mPFS) of 24 weeks [w] and clinical benefit rate (complete/partial response or stable disease ≥ 12 weeks) of 53% on matched tx. Comparison by NGS platform showed improved mPFS for matched vs unmatched pts sequenced on FM1 vs IAC (FM1: 26 vs 19w, HR = 0.76 [95% CI: 0.3-1.9]; IAC: 8 vs 12w; HR = 1.21 [95% CI: 0.5-2.8]). Interestingly, 1 pt with SMARCB mutation, reportedly associated with the FGFR pathway, had a PFS of 70w on tx with a pan-FGFR inhibitor after progressing on 3 prior lines of tx (ET and CT). Conclusions: Molecular profiling of MBC in a phase I unit led to matched tx in 25% of cases. Matched pts showed encouraging mPFS with a suggestion of benefit in those matched after sequencing on a broader gene panel (FM1).

4591 Background: Cholangiocarcinoma (CCA) is the most common biliary tract malignancy with an estimated incidence of 8,000–10,000 patients/year in the US. Chemotherapy is the most common second-line treatment with reported outcomes in patients with CCA. Response rates of < 10% and median progression-free survival (PFS) times of ~3–4 months have been reported with second-line chemotherapy regimens, including FOLFOX in the ABC-06 trial. Fibroblast growth factor receptor 2 ( FGFR2) fusions occur in 13–17% of CCA and multiple targeted agents are in development for patients with FGFR2 fusions. To date, the outcome of patients with CCA and FGFR2 fusions receiving standard second-line chemotherapy is unknown. Methods: Patients with advanced CCA and FGFR2 fusions after prior treatment with gemcitabine-based chemotherapy were enrolled in a single-arm phase 2 study (NCT02150967) and received the FGFR1–3 selective TKI infigratinib (previously BGJ398) 125 mg orally qd on d1–21, cycles repeated q28 days until unacceptable toxicity, disease progression, investigator discretion, or withdrawal of consent. A retrospective analysis of a subset of patients who received infigratinib as third- or later-line treatment was performed. Investigator-assessed PFS and best overall response (BOR, per RECIST 1.1) following second-line chemotherapy (pre-infigratinib) and third-line or later-line infigratinib were calculated. Results: Of the 71 patients (44 women; median age 53 years) with FGFR2 fusions enrolled at the time of analysis (datacut 8 August 2018), 37 (52%) were included in this retrospective analysis. Median PFS with standard second-line chemotherapy was 4.63 months (95% CI 2.69–7.16) compared with 6.77 months (95% CI 3.94–7.79) for third- and later-line infigratinib. BOR for second-line chemotherapy was 5.4% (95% CI 0.7–18.2) compared with 21.6% for third- and later-line infigratinib (95% CI 9.8–38.2). Conclusions: Outcomes from second-line chemotherapy in patients with CCA and FGFR2 fusions were similar to those reported in the literature for all patients with CCA regardless of genomic status and remain dismal. Infigratinib administered as third- and later-line treatment resulted in a meaningful PFS and ORR benefit in patients with CCA and FGFR2 fusions. Clinical trial information: NCT02150967 .

4508 Background: The combination of dual immune checkpoint inhibitors (ICI) T (anti–CTLA-4) and D (anti–PD-L1) showed tolerability with a promising objective response rate (ORR) in the initial cohort of this study (NCT02519348). Subsequent evaluation of pts with solid tumors treated with increasing doses of T suggested priming with a higher dose of T may induce a stronger immune response and enhance anti-tumor activity. Thus, the randomized expansion cohorts comprised 4 arms evaluating T and D as monotherapies and 2 T+D regimens, including a novel T+D regimen featuring a single, priming dose of T. Methods: ICI-naïve pts with aHCC who progressed on, were intolerant to, or refused sorafenib were randomized to one of two T+D combinations: T300+D (T 300 mg + D 1500 mg 1 dose followed by D Q4 weekly [Q4W]) or T75+D (T 75 mg Q4W + D 1500 mg Q4W [4 doses] followed by D Q4W); or single agent D (1500 mg Q4W) or T (750 mg Q4W). Safety was the primary endpoint. ORR by blinded, independent central review using RECIST v1.1, duration of response (DoR), circulating lymphocytes, and overall survival (OS) were assessed. Results: At data cut-off (09/02/2019), 332 pts were enrolled. Median follow-ups were 11.7 months (mo) for T300+D, 14.6 (T75+D), 8.9 (D), and 15.8 (T). Treatment-related adverse event (trAE) incidences are shown (Table); no deaths were attributed to trAEs for T300+D or T. The T300+D arm had the highest confirmed ORR (DoR not reached [NR]) and longest OS (Table). A unique proliferative T cell profile was identified for pts in the T300+D arm, suggesting additive biologic activity for the combination, and showed pts with an OR exhibited high cytotoxic (CD8) counts. Conclusions: The encouraging clinical activity and tolerable safety profile suggest T300+D provides the best benefit-risk profile as opposed to T75+D or monotherapies. The unique pharmacodynamic activity of the T300+D regimen further supports its use in aHCC. T300+D and D are being evaluated in the ongoing phase III HIMALAYA study (NCT03298451) in first-line HCC vs sorafenib. Funding: AstraZeneca. Clinical trial information: NCT02519348 . [Table: see text]

e16623 Background: The management of aHCC has evolved dramatically in recent years, with new agents like immunotherapy receiving regulatory approval. As we begin incorporating these drugs into routine clinical practice, data on real-world sequencing of therapies and clinical outcomes are needed. Methods: We utilised the DISCOVERYAI platform, a virtual machine containing de-identified patient electronic health records to review HCC patients treated at the National University Health System, Singapore from January 2015 to December 2019. We then identified those who received systemic therapy and correlated their clinical outcomes. Results: In total, 395 HCC patients were identified; 75 received surgery, 174 received loco-regional therapy and 102 referred for consideration of systemic therapy. Of those considered for systemic therapy, median age was 65 years (range 23-87); 88% male (n = 90); hepatitis B/hepatitis C/non-hepatitis, 41(40.2%)/ 10(9.8%)/ 51(50.0%). 75.5% (n = 77) of them received systemic therapy with a TKI and/or immunotherapy. 39% (n = 30) of these received second-line treatment. Child-Pugh score at start of treatment was A5/A6/B7/B8, 38(49.3%)/ 32(41.6%)/ 5(6.5%)/ 2(2.6%) respectively. In the first-line, 66% (n = 51) received TKI and 34% (n = 26) received immunotherapy. Amongst those treated with first-line TKI, 45% (n = 23) received second-line therapy; 65% (n = 15) immunotherapy, 35% (n = 8) another TKI. Of those treated with first-line immunotherapy, 27% (n = 7) received second-line TKI. At a median follow-up of 35 months, first-line median progression-free survival (mPFS) for TKI vs immunotherapy was 3.7 vs 3.1 months (HR 0.73; 95% CI, 0.40-1.33; p = 0.31). mPFS for second-line immunotherapy vs TKI was 4.0 vs 2.9 months (HR 0.43; 95% CI, 0.19-0.96; p = 0.04). When comparing sequencing of therapies, the combined first and second mPFS for TKI-immunotherapy/TKI-TKI/immunotherapy-TKI is 9.5/7.6/7.6 months respectively (log-rank test, p = 0.71). Those patients that received both immunotherapy and TKI had significantly higher overall survival (OS) compared to those receiving only immunotherapy or only TKI or none (mOS NR vs 10.1 vs 13.2 vs 4.7 months; p < 0.001). Conclusions: TKI remains an important pillar of treatment in aHCC in the era of immunotherapy. While immunotherapy provides long durable responses and benefit in a minority of patients, the majority appear to benefit from TKI. Biomarker studies are needed to discern treatment algorithms for aHCC.

e19385 Background: Pembrolizumab has dramatically improved the survival of patients with non-small cell lung cancer (NSCLC) and is considered the standard of care for first line treatment of NSCLC who do not harbour oncogenic drivers. The fixed dose of 200mg was approved by the US Food and Drug Administration. The dose of 200mg was based on pharmacokinetic analysis. Studies have demonstrated equivalent efficacy with weight-based dosing 2mg/kg. An average Asian weighs 50-60kg. We aimed to look at the efficacy of pembrolizumab at a low fixed dose compared to the standard dosing. Methods: A review of all consecutive patients receiving pembrolizumab for advanced NSCLC from January 2016 to December 2019 in a large, high-volume academic medical centre, the National University Hospital, Singapore was conducted. Data fields collected include patient’s demographics, treatment doses and clinical characteristics. Time on treatment and overall survival were analysed using the Kaplan Meier method. Results: In total, 92 ECOG 0-2 patients with advanced NSCLC were treated with pembrolizumab. Median age was 69 years (Range, 29-92). Most were males (76%) and Chinese race (68%). Of the 92 patients, 46 (50%) and 46 (50%) received 100mg (Pem100) and 200mg (Pem200) of pembrolizumab respectively. Pembrolizumab was prescribed as first line in 73 (79%) and second line in 19 patients (21%). The average dose of pembrolizumab received in the low dose group was 1.87mg/kg (Range, 1.24mg/kg – 2.70mg/kg). 88 patients were included in the survival analysis. 4 were excluded due to the presence of an oncogenic driver. Patients were followed up for a median of 13.2 months. There was no difference in progression free survival between Pem100 and Pem200 for first-line single agent and when combined with chemotherapy (PFS: NR versus 5.3months, HR 2.17, 95% CI 0.76-6.16, p = 0.15 and NR vs 16.9 months, HR 2.89, 95% CI 0.35-25.16, p = 0.33 respectively). For patients who received pembrolizumab in the first line setting, the response rate was 56% vs 20% (p = 0.07), 67% vs 52% (p = 0.69) for Pem100 and Pem200 as a single agent and when combined with chemotherapy respectively. The median number of cycles received was 8.9 (Range, 1-60 cycles), translating to estimated cost savings of SGD 45 395 (~ USD 32 664) per patient who received Pem100. Conclusions: A lower fixed dose of pembrolizumab at 100mg showed no difference in progression free survival and response rate in an Asian cohort with significant cost savings. A further randomised controlled trial in an Asian population should be carried out.

4597 Background: Patients with advanced or metastatic BTC who progress on first-line (1L) gemcitabine-based doublet chemotherapy have few 2L treatment options. Varlitinib is a reversible small molecule pan human epidermal growth factor receptor (HER) inhibitor with low nanomolar potency against HER1 (EGFR), HER2 and HER4 with promising early results in advanced BTC. Methods: TreeTopp (NCT03093870) was a global, multicenter, double blind phase 2 study in which patients with advanced BTC who progressed after 1L therapy that included ≥6 doses of gemcitabine, with radiographically measurable disease based on RECIST v1.1, ECOG PS 0 or 1 and albumin ≥3 g/dL were randomized (1:1) to varlitinib (300 mg BID) plus capecitabine (1000 mg/m2 BID 14 days on/ 7 off)(V+C) or placebo plus capecitabine (P+C). The dual primary endpoints were Objective Response Rate (ORR) and Progression Free Survival (PFS) defined as the time from randomization to radiological progression assessed by Independent Central Review. Secondary end points included Overall Survival (OS). Results: Overall, 127 patients were randomized (V+C, n = 64; P+C, n = 63) from May−Dec 2018 and demographics/baseline characteristics were generally well balanced, although the V+C arm had a lower proportion of females vs. P+C (31% vs. 48%). The odds ratio for ORR was numerically higher with V+C vs. P+C was 2.278 (9.4% vs. 4.8%, p = 0.42), the HR for PFS for V+C vs. P+C was 0.90 (median PFS, 2.8 vs. 2.8 months; p = 0.63), and the HR for OS for P+C vs. V+C was 1.11 (median OS, 7.8 vs. 7.5 months; p = 0.66). Although not powered to evaluate sub-group interactions, in sub-group analysis, V+C showed PFS benefit versus P+C in two sub-groups; gallbladder cancer (GBC, HR = 0.55, 95% CI: 0.25, 1.22; median PFS, 2.9 vs. 1.6 months) and females (HR = 0.59, 95% CI: 0.28, 1.23; median PFS, 4.1 vs. 2.8 months). There was no PFS benefit for V+C vs. P+C among males and non-GBC. Toxicities were generally balanced between arms apart from a slightly higher incidence of hyperbilirubinemia, diarrhea and fatigue in the V+C vs. P+C arm. Grade 3/4 toxicities were reported in 66% and 59% of patients in the V+C and P+C arms, respectively. Conclusions: V+C is well tolerated but did not improve ORR, PFS or OS vs. P+C in 2L advanced BTC. Exploratory analyses suggested that patients with GBC and female patients achieved comparatively higher median PFS with V+C vs. P+C. Clinical trial information: NCT03093870 .

360 Background: PIPAC is a novel, laparoscopic intraperitoneal chemotherapy delivery technique which aims to improve on hyperthermic intraperitoneal chemotherapy (HIPEC), ameliorating drug distribution and tissue penetration. Thus far, PIPAC has been conducted with oxaliplatin chemotherapy in Europe, at an arbitrary dose of 92mg/m2; 150mg/m2 was found to be intolerable. We conducted a dose-escalation phase 1 study to establish the safety, tolerability and recommended phase 2 dose (RP2D) for PIPAC in Asian patients. Methods: This phase 1 study of oxaliplatin administered via PIPAC was designed as a traditional 3+3 dose escalation study for patients with predominant peritoneal metastasis from a gastrointestinal primary tumor, after failure of standard therapies. Dose levels were planned at 45, 60, 90 and 120mg/m2. Repeat doses of PIPAC were permitted, 6 weeks apart. Dose limiting toxicities (DLT) were defined as any clinically relevant grade 3 adverse events occurring within 28 days after PIPAC. Results: This study included 16 patients (25 PIPAC procedures; 8 gastric, 4 colorectal and 1 gallbladder, pancreas and appendix cancer each). Median age was 62 years, with a median peritoneal carcinomatosis index (PCI) score of 17 (range 1 - 39). Two patients developed pancreatitis (grade 2 and 3) on day 6 and day 9 after PIPAC administration at the dose cohort of 45mg/m2, necessitating cohort expansion to 6 patients. One patient was noted to have asymptomatic grade 3 hyperamylasemia (90mg/m2 cohort). There were no other DLTs and all 3 patients in the highest dose cohort (120mg/m2) tolerated PIPAC well. Nine patients who underwent a 2nd PIPAC procedure had a decrease in PCI score from 18.4 to 15.5; one patient at 120mg/m2 had an improvement in PCI from 30 to 12. Conclusions: The RP2D of PIPAC with oxaliplatin is 120mg/m2. Single agent PIPAC is well tolerated, and future studies with PIPAC must consider a bi-directional approach with the incorporation of systemic therapy, with either chemotherapy or immunotherapy to improve efficacy. Clinical trial information: NCT03172416.

4095 Background: FGF19 overexpression is hypothesized to hyperactivate FGFR4 and its downstream signaling pathway leading to enhanced tumor growth in HCC/ICC. Targeting FGFR4 may have therapeutic benefit in HCC/ICC with altered FGF19 signaling. A phase 1 study (NCT02834780) was initiated to assess H3B-6527, an investigational highly selective covalent FGFR4 inhibitor. Methods: Adult pts with advanced HCC or ICC, ECOG PS 0-1, well compensated liver function, and who progressed after at least one prior therapy, were administered H3B-6527 orally QD (once daily) on a 21-day cycle following a 3+3 design. Patients in the dose escalation phase were treated regardless of FGF19 status. Adverse events (AEs), pharmacokinetics (PK), and pharmacodynamics (PD) were assessed. Response was determined by RECIST 1.1 or modified RECIST every 6 weeks. Results: As of 06-Jan-2019, 37 pts have been treated with H3B-6527 at doses of 300 to 1400 mg QD (23 pts in escalation; 14 in expansion). In dose escalation, a total of 17 patients with HCC, Child-Pugh A received prior systemic therapy including 100% with prior TKI and 35% with prior IO. 12% had hepatitis B virus and 47% had hepatitis C virus. H3B-6527 plasma levels increased with dose from 300 to 1000 mg QD and plateaued. H3B-6527 was rapidly absorbed with a tmax of ~2-3 h and showed a terminal half-life of ~4-5 h, following administration of 1000 mg (fasted). No dose-limiting toxicities or ≥ Grade 3 treatment-related AEs (TRAE) have been observed in escalation. Most common TRAEs (≥ 10%) were diarrhea, nausea, and vomiting. Based on safety, PK, and PD, 1000 mg QD was the recommended phase 2 dose. Durable stable disease and partial responses (PR) have been observed on the once daily fasted schedule; 2 of 17 pts with HCC achieved PRs and an additional 7 with stable disease were on treatment for ≥ 5 months. Conclusions: H3B-6527 is well tolerated and demonstrates early signs of clinical activity. Dose expansion on QD schedule and exploration of BID (twice daily) schedule is ongoing. Clinical trial information: NCT02834780.

1045 Background: Endocrine blockade (EB) is standard of care for patients (pts) with HR+ LABC/MBC. RET over-expression (RET+) occurs in up to 75% of HR+ breast cancers and is a postulated mechanism of endocrine resistance. Preclinical studies show cross talk between RET and estrogen receptor, and at least additive treatment (Tx) effect of Len+EB. Methods: We performed a phase Ib trial (3+3 dose escalation) to study safety, tolerability, pharmacodynamics and efficacy of Len+Let. Both drugs were given as continuous daily dosing with 2 weeks (wks) of Len alone, followed by Len+Let for 12 wks then surgery (LABC), or till disease progression (PD) (MBC). Serial tumor biopsies (n = 15) were done at baseline, after Len alone, 4 wks post Len+Let, and at surgery [LABC] / upon PD [MBC]. Results: 16 pts were treated (4 LABC, 12 MBC); Among MBC pts, median lines of prior Tx was 3 (range 0-10); 84.6%, 66.7%, and 58.3% had prior EB, EB+CDK4/6 inhibitor (i), and chemotherapy (CT) respectively. At dose level (DL) 1, 2/4 pts had dose-limiting toxicities (DLT). There was no DLT at DL-1, but 6/6 pts needed dose reductions (DR), with 4/6 DR within 6 wks of Len+Let (3 G3 hypertension [HTN], 1 G3 wound pain), deeming DL-1 intolerable. At DL-2, 0/6 pts had DLT; this was declared recommended phase 2 dose (RP2D). Most frequent G3 toxicities (tox) were HTN (6/16), proteinuria (2/16) and palmar-plantar erythrodysesthesia (PPE) (2/16), with no G4/5 tox. Len+Let was active with 93.8% overall disease control rate (DCR) (50.0% partial response [PR], 43.8% stable disease [SD]). Among MBCts (8/12 had prior EB+CDK4/6i), DCR ≥12 wks was 91.7%; 1 pt had sustained PR for 48 wks and 1 ongoing PR at 40 wks. 9/16 pts had RET+ tumors on immunohistochemistry at baseline, and 66.7% showed down-regulation with Tx (RECIST: 4 PR, 2 SD). Conclusions: Len+Let showed significant anti-tumor activity, even in pts who failed prior CT or EB+CDK4/6i. RP2D of 14mg Len and 2.5mg Let is tolerated with efficacy; dose expansion is currently underway. Clinical trial information: NCT02562118. [Table: see text]

4065 Background: High mortality from gastric cancer is related to the late manifestation of its symptoms. A blood-based non-invasive biomarker with the ability to detect all stages of gastric cancer could significantly improve patient outcomes. We aimed to develop a novel serum miRNA assay for diagnosis of gastric cancer. Methods: We conducted a multi-center study involving 892 gastric cancer and control subjects from Singapore and Korea to develop a multi-target miRNA assay. Using RT-qPCR, we quantified the expressions of 578 serum miRNAs and constructed a 12-miR biomarker panel through multi-variant data analysis. The results were generated with the use of a logistic-regression algorithm, with the value of 40 or more considered to be positive. We subsequently validated this multi-miR assay in a large prospective cohort involving 4566 subjects and compared its performance with traditional markers such as H.Pylori and Pepsinogen. All participants underwent gastroscopy independent of the assay results. Results: Of the 4566 subjects that underwent gastroscopy and histopathological examination in the prospective cohort, 125 were diagnosed with gastric cancer. The 12-miR assay achieved an Area-Under-Curve (AUC) of 0.84, significantly outperforming (p-value < 0.01) that of H.Pylori (AUC of 0.64) and Pepsinogen (AUC of 0.62). The sensitivity of the miRNA assay in detecting early (stage 0-2) and late (stage 3-4) stage gastric cancer was 82.6% (95% CI, 68.6% to 92.2%) and 88.4% (95% CI, 78.4% to 94.9%) respectively at a specificity of 70.0% (95% CI, 67.8% to 71.9%). In comparison, H.Pylori showed a sensitivity of 80.4% at a specificity of 44.3% whereas the Pepsinogen showed sensitivity of 9.52% at a specificity of 95.3%. Using the miRNA assay as a pre-screening tool could potentially reduce number of endoscopy needed by 62% in detecting one case of gastric cancer. Conclusions: Our serum miRNA panel is a useful, non-invasive screening test for gastric cancer. It is cost-effective as it can reduce unnecessary diagnostic endoscopy.

319 Background: BTC is a rapidly progressing cancer with limited response to chemotherapy. First line therapy (gem/cis) was defined in the ABC-02 study (Valle et al), where a 26% response was seen. HER family receptors are overexpressed and may be involved in tumour proliferation and survival in BTC. Varlitinib, a reversible tyrosine kinase inhibitor of EGFR, HER2, and HER4, shows potent activity as monotherapy in preclinical BTC models and clinical activity in BTC patients (pts) in phase (ph) 1 trials. We conducted a ph Ib/2 study of varlitinib plus gem/cis in BTC to understand the safety profile and determine the maximum tolerated dose (MTD) of the combination. Methods: A modified 3+3+3 escalation design was used in Ph1b, with 2 varlitinib dose levels (200 and 300 mg BID) plus gem/cis on Day 1 and 8 in a 3 week cycle. The primary objectives are to determine the MTD and to characterize the safety profile. Secondary objectives are to assess the preliminary efficacy and to evaluate the pharmacokinetics of varlitinib and any circulating metabolites. Results: As of 10 Sep 2018, 21 pts were enrolled (11 in 200 mg cohort and 10 in 300 mg cohort, with 9 and 4 evaluable for MTD, respectively). Dose limiting toxicities (DLTs) were observed in 3 pts (1 G3 unconjugated hyperbilirubinemia and 1 G3 ALT transaminitis/G4 AST transaminitis in 200 mg cohort; 1 G4 thrombocytopenia/G3 febrile neutropenia/G3 AST elevation in 300 mg cohort). In 19 pts who were on varlitinib ≥ 1 month, 7 had partial response and 10 achieved stable disease (all > 12 weeks), giving the overall response rate of 37% and the disease control rate of 89%. The median PFS for the 200 mg cohort was 248 days and was not reached for the 300 mg cohort. The most common (≥ 30%) all-grade adverse events (AEs) regardless of causality were thrombocytopenia (62%), neutropenia (52%), anorexia (38%), nausea (38%), diarrhoea (38%), and anaemia (33%); the most common (≥ 15%) grade ≥ 3 AEs were neutropenia (48%), thrombocytopenia (33%), and anaemia (19%). Conclusions: Varlitinib plus gem/cis was well tolerated in the 200 mg cohort; the 300 mg cohort is ongoing. Preliminary anti-tumour activity was observed. Data will be updated at the time of presentation. Clinical trial information: NCT02992340.

29 Background: The incidence rates of gastric cancer in Asia are among the highest in the world. Gastric cancer is a heterogeneous disease. The Cancer Genome Atlas and Asian Cancer Research Group provide molecular classification of gastric cancers. Recently, immunotherapy was found to be effective in selected gastric cancer patients. Despite the molecular classification, a classification system based on immune microenvironment is in need to better stratify patients for ideal immunotherapy combinations. Methods: We have performed indexed-sorting single cell RNA-seq to analyze both CD45+ and CD45- cells from biopsy samples of 15 gastric cancer patients (5 Diffuse; 4 Intestinal; 5 Mixed; 1 unknown type). Using our ezsinglecell analytical pipeline, we have built a draft Asian Single-cell Immune Atlas (ASIA) of human gastric cancers. Results: We were able to probe fibroblast, epithelial cells and various subtypes of immune cells. Interestingly, we observed that fibroblast and mast cells were more enriched in diffuse type. In addition, we have profiled single immune cells in tumor and normal adjacent tissues to distinguish tissue-induced versus tumor-driven immune signatures. Comparing tumor and normal tissues, we found that tumor site has less natural killer (NK) and CD8 T cells; more B, myeloid and CD4 T cells. Moreover, CD8 T cells and NK cells showed distinct phenotypes in tumor compared to normal tissues. In particular, inhibitory receptor Tim-3 and CTLA-4 were up-regulated in the tumor. Most interestingly, T Cell Receptor (TCR) reconstruction revealed clonal expansion of CD8 T cells. Conclusions: Both immune (CD45+) and non-immune (CD45-) cell components of human gastric cancers are highly heterogeneous. Diffuse type showed differential tumor microenvironment compared to intestinal and mixed types. Tumor site presented an immune-suppressive phenotype compared to normal adjacent site. The clonal expansion of CD8 T cells suggested the role of adaptive immunity on recognizing tumor antigens. This discovery may determine the clinical applications of immunotherapy for gastric cancer in the future.

Purpose Biliary cancers overexpress epidermal growth factor receptor (EGFR), and angiogenesis has been correlated with poor outcome. Erlotinib, an EGFR tyrosine kinase inhibitor, and bevacizumab, a vascular endothelial growth factor (VEGF) inhibitor have each been shown to have activity in biliary cancer. The primary objective of this study was to evaluate the response rate by Response Evaluation Criteria in Solid Tumors (RECIST). Secondary end points included overall survival (OS), time to progression (TTP), VEGF levels, and molecular studies of EGFR and k-ras. Patients and Methods Eligible patients had advanced cholangiocarcinoma or gallbladder cancer. Patients were treated with bevacizumab 5 mg/kg intravenously on days 1 and 15 and erlotinib 150 mg by mouth daily on days 1 through 28. Responses were evaluated by RECIST. VEGF levels were collected, and samples were analyzed for EGFR mutation by polymerase chain reaction. Results Fifty-three eligible patients were enrolled at eight sites. Of 49 evaluable patients, six (12%; 95% CI, 6% to 27%) had a confirmed partial response. Stable disease was documented in another 25 patients (51%). Rash was the most common grade 3 toxicity. Four patients had grade 4 toxicities. Median OS was 9.9 months, and TTP was 4.4 months. Low repeats (< 16) in EGFR intron 1 polymorphism and G>G k-ras Q38 genotype (wild type) were associated with improved outcomes. Conclusion Combination chemotherapy with bevacizumab and erlotinib showed clinical activity with infrequent grade 3 and 4 adverse effects in patients with advanced biliary cancers. On the basis of preliminary molecular analysis, presence of a k-ras mutation may alter erlotinib efficacy. The combination of bevacizumab and erlotinib may be a therapeutic alternative in patients with advanced biliary cancer.

2588Background: Varlitinib (V) is a reversible inhibitor of HER1/HER2/HER4. This is a phase 1b dose confirmation study to determine safety and early efficacy signals of V ± T combined with weekly P...

e18748Background: Patient selection for phase I (P1) trials in oncology is challenging. A typical inclusion criterion for P1 studies is 'life expectancy > 3 months', however overall survival (OS) o...

2550 Background: Genomic profiling of single core biopsies (bx) are confounded intratumoral heterogeneity, resulting in sampling bias. We explored the use of a novel technique to obtain multiple bx from single metastatic lesion in pts to evaluate heterogeneity and identify therapeutic ATM. Methods: 15 pts (5 NSCLC; 3 ovarian; 2 colon, 2 uterine and 1 breast, cervix and HCC) with biopsiable lesions were identified. Using a single pass radiologically guided percutaneous bx technique, we obtained multiple spatially distinct core bx samples from a single metastatic lesion. Each bx underwent DNA extraction and WES using the NextSeq500. Results: Median of 4 core bx were obtained from each lesion. Complication rate utilizing this technique was 0%. 2 pts were omitted from analysis due to poor quality DNA with 13 pts successfully sequenced. In 1 pt, only 2 of 4 cores were successfully sequenced. The median amounts of total and non-synonymous variants were 137 (27-1286) and 66 (10-649) respectively. The median (range) filtered variants detected in 1/4, 2/4, 3/4, and 4/4 bx cores was 63(16-91)%, 5(1-65)%, 4(0-30)% and 26(0-63)% respectively, suggesting significant subclonal diversity within a single lesion. ATMs were identitified in 8/13 pts. 4/13 pts (31%) had no ATM across all 4 cores. 3 pts received therapy with inhibitors targeting ATMs. A pt with AKT1_E17K ATM received an AKT inhibitor with 21% tumour shrinkage and PFS 6.1 mths. 2 NSCLC pts harbouring an EGFR_T790M ATM were treated with an EGFRT790M specific TKI. 1 withdrew due to toxicity after 2mths and another had PFS > 16.5 mths. Tumour mutational burden (TMB) was consistent across multiple bx from each lesion. A NSCLC pt with the highest TMB received a checkpoint inhibitor with ongoing > 4 mths stable disease. Conclusions: Utilizing a single pass radiologically guided techniqueto obtainmultiple bx is feasible, safe and informative. This allows reconstruction of a tumour’s subclonal genomic architecture, providing insights into mutational heterogeneity and help guide therapy.

4073 Background: Durvalumab and tremelimumab, investigational monoclonal antibodies against PD-L1 and CTLA-4 immune checkpoints, respectively, have shown efficacy in monotherapy and offer promise in combination for patients (pts) with HCC. This is a phase I/II, open-label, randomized study of durvalumab combined with tremelimumab in unresectable HCC. Methods: Phase I part of this study is a safety run-in cohort treated at the recommended phase II doses of the durvalumab/tremelimumab combination (20 and 1 mg/kg IV Q4W respectively for 4 doses followed by 20 mg/kg Q4W durvalumab alone) in pts with unresectable HCC with or without concomitant HBV or HCV infection who progress on, are intolerant to, or have refused sorafenib therapy. Secondary objectives include evaluation of antitumor activity. Here we present results of a preplanned analysis from the completed phase I part of the study. Results: As of 10 January 2017, 40 pts have been enrolled (11 HBV+, 9 HCV+, 20 uninfected). 30% had no prior systemic therapy; 93% were Child Pugh Class A. 24 (60%) had ≥1 treatment-related AE; 20% had ≥1 grade ≥3 related AE. Most common (≥15%) treatment-related AEs: fatigue (20%), increased ALT (18%), pruritus (18%), and increased AST (15%). Most common grade ≥3 related AE was asymptomatic increased AST (10%). 24 pts have discontinued treatment: 3 due to treatment-related AEs (grade 3 pneumonitis, grade 3 colitis/diarrhea, asymptomatic grade 4 elevated AST and ALT), 16 due to progressive disease, 4 due to death unrelated to treatment (cardiac arrest, variceal bleed, progressive disease, probable HCC rupture), and 1 other (pt entered hospice care). 40 pts were evaluable for response at ≥16 weeks follow-up. Conclusions: No unexpected safety signals with durvalumab and tremelimumab were seen in this unresectable HCC population. Clinical activity observed predominantly in uninfected pts though interpretation limited by small subsets. Enrollment to the phase II portion of the study is ongoing. Clinical trial information: NCT02519348. [Table: see text]

176 Background: AA is a standard of care for the treatment of metastatic CRPC, with an approximate retail cost of $8000/month. Despite a large food effect (~17-fold increase in CMax with a high-fat meal), AA was administered under fasting conditions in its pivotal trials. There are no randomized continuous dosing studies examining prandial state dosing on AA PK/PD. We sought to test the hypothesis that LOW (250mg w/food) would have similar PK/PD and safety to standard (STD, 1000mg fasting) in patients with progressive CRPC. Methods: Patients (n = 72) with progressive CRPC from seven institutions in the US and Singapore were randomized to treatment with STD or LOW (with low-fat breakfast). Both arms received prednisone 5mg twice daily. PSA was assessed monthly, and testosterone, DHEA/DHEAS were assessed every 12 weeks along with standard disease burden assessments. PK samples were collected at day 1, 8 and months 2, 3, 4. Log change in PSA response rate from baseline to week 12 was examined as the primary endpoint, with a non-inferiority design based on a non-inferiority margin of 15%. This margin corresponds to a 0.51 standard deviation (SD) difference in the mean log changes between the groups. Results: Accrual was completed with n = 36 on STD and n = 36 on LOW. Median (range) age was 74 (52-89) and baseline PSA was 39.2(range 0.6-1789). Mean log-change in PSA at 12 weeks was nominally greater in the LOW arm (-1.59 vs. -1.19). The 95% confidence for the difference (STD-LOW) ranged from -0.40 to 1.19, with lower limit corresponding to -0.24SD. Thus, non-inferiority of LOW was established. Median time to PSA progression was ~14 month in both arms (p = 0.53). Preliminary analysis of PK showed no difference in CMaxbeyond the first cycle with lower PK variability in the LOW arm. Conclusions: Low-dose (250mg/day) abiraterone acetate with a low-fat breakfast is non-inferior to standard dosing in a fasted state with respect to PSA and PK metrics. Although PSA response and progression are not clinically validated surrogates, given the pharmacoeconomic implications, these data warrant consideration by prescribers and payors. Clinical trial information: NCT01543776.

48 Background: Theoxaliplatin/S-1 (SOX) and cisplatin/S-1 (SP) regimens are interchangeably used in the management of advanced gastric cancer (AGC). We previously developed a classification tool using gene expression signatures which successfully stratified gastric cancer cell lines and primary tumour samples according to their sensitivity to SOX or SP ( Tan et al, Gastroenterology 2011). We now report the first prospective study to evaluate the feasibility and efficacy of using a genomic classifier to tailor treatment in this setting. Methods: Pts with histologically-confirmed locally-advanced, metastatic and recurrent GC were recruited from 3 centres in Singapore and South Korea. Tumours were analysed using the Affymetrix HG-U133 Plus 2.0 array and results were used to classify pts as G1 (oxaliplatin-sensitive), G2 (cisplatin-sensitive) and G3 (status unclear or gene expression not available). G1 and G2 pts were matched to SOX and SP regimens respectively, while G3 pts were assigned SOX. The primary endpoint was best overall response; secondary endpoints were turnaround time and biomarker analyses. Results: Between July 2, 2010, and Apr 2, 2015, we screened 85 AGC pts. 74 pts received at least 1 cycle of treatment and were evaluable for analysis. Median turnaround time was 7 working days (IQR, 5–9). The misclassification rate was 6%. After an initial 30 pts in the G1 subgroup were treated with SOX, subsequent pts ( N = 13) classified as G1 received the SP regimen. The ORR were 44.8%, 8.3%, 26.7% and 55.6% for G1 SOX, G1 SP, G2 SP ( N = 19), G3 SOX ( N = 12) respectively; and was higher in G1 pts treated with SOX compared with SP ( P = 0.033). Post hoc genomic reclassification based on Lei et al (Gastroenterology 2013) confirmed the utility of the metabolic subtype as a predictive marker of benefit from chemotherapy (log rank P value for PFS = 0.004). Conclusions: This bench-to-bedside effort establishes that molecular profiling to direct choice of conventional chemotherapy for AGC is possible within a reasonable timeframe. The clinical utility of our genomic classifiers in question are promising but warrant further investigation. Clinical trial information: NCT01100801.

e14124Background: Low dose, short course anti-angiogenic agents may normalize tumor vasculature to improve drug delivery. We previously showed 12.5mg sunitinib (Su) for 7 days before neoadjuvant do...

4060 Background: Gastric cancer is the 2nd most common cause of cancer deaths worldwide. Currently, endoscopy is the only reliable method for early diagnosis. However, the invasiveness and cost lim...

2542 Background: SXR is an orally administered potent XPO1 inhibitor that forces nuclear retention and activation of multiple tumor suppressor proteins resulting in tumor cell death. SXR was evalua...

65 Background: The aim of this study was to evaluate the safety and optimal dosing schedule of a HLA-restricted cancer vaccine cocktail, OTSGC-A24 targeting novel specific tumor antigens FOXM1, DEPDC1, KIF20A, URLC10 and VEGFR1 in advanced gastric cancer patients with HLA-2402 haplotype. Methods: The vaccine was administered subcutaneously at 3-weekly, 2-weekly and weekly interval. In the first phase, 3 evaluable patients were treated in each dose cohort. The decision to expand the 2-weekly and weekly cohort to 10 patients per cohort was based on specific cytotoxic T-lymphocyte (CTL) induction rate using ELISPOT assay 4 weeks after vaccination. Results: The vaccine was well tolerated at all dose cohorts. The most common toxicity seen was skin induration at the injection sites. No DLT was observed. One patient developed G2 interstitial pneumonitis, 6 months after initiation of treatment. Stable disease was seen in 4 of 9 evaluable patients. The specific CTL induction can be observed at 4 weeks post vaccination with the highest (>2 of 3) observed in the weekly and 2-weekly cohorts. One of 3 patients in the 3-weekly cohort achieved specific CTL response at 4 weeks post vaccination, although all patients attained specific CTL response by 12 weeks after vaccination. Conclusions: OTSGC-A24 combined peptide vaccine was well tolerated. At 4 weeks post vaccination, the CTL induction rate is superior in the weekly and 2-weekly cohort. Efficacy data of the expanded cohorts will be presented. Clinical trial information: NCT01227772.

127 Background: The addition of trastuzumab to cisplatin and 5-fluorouracil or capecitabine has been shown to have superior survival benefit. In this study, the addition of trastuzumab to chemotherapy combination, S-1 (tegafur, gimeracil, oteracil potassium; TS-ONER) and cisplatin, a standard first line therapy for advanced gastric cancer in Japan, was evaluated for efficacy and safety. Methods: This was a single-arm, phase II trial recruiting patients with advanced HER-2 positive (IHC 3+ or FISH +) gastric adenocarcinoma who had not had prior systemic therapy. Patients included had measurable disease, ECOG 0-2, creatinine clearance >60ml/min, LVEF at least 50%. Patients accrued received cisplatin at 60mg/m2 on day 1 of each cycle, trastuzumab at 8mg/kg as loading dose followed by 6mg/kg on day 1 of each cycle and oral S-1 for 14 days ( dosing based on BSA), every 3 weeks. Treatment continued till progressive disease or unbearable toxicity. Primary endpoint was response rate (ORR) and secondary endpoints were PFS, OS, CBR and safety. The study was designed to distinguish a favorable ORR of 55% from a null rate of 30% at 5% significance level and 80% power. Results: A total of 30 patients were recruited. Median age was 61.9 years, 73% were males and 90% had no prior resection of primary tumor. 56.7% were Chinese, 23.3% Japanese and 20% Korean. One patient achieved complete response, 16 had partial response and 2 had stable disease >24 weeks (ORR 63%; CBR 70.4%). Three patients did not have evaluable disease. With a median follow-up of 7.2 months, the median PFS was 7.4 months with 67.7% free of disease progression at the 6-month mark. Median OS was 14.6 months. The most common AEs were anorexia (50%), fatigue (47%), diarrhea ( 47%), nausea (37%), neutropenia (33%), anemia (27%) and mucositis ( 27%). LVEF decreased

2551 Background: Thymidylate synthase (TYMS) is the target of fluoropyrimidines (FP). TYMS 3R3R is the predominant genotype in East Asian (EA) patients and is associated with increased TYMS expression, reduced FP related toxicity and relative FP resistance. Dose finding studies for FP were developed in Caucasians, a population that typically harbors TYMS 2R2R and 2R3R genotypes. We hypothesize the recommended phase II dose (RP2D) of capecitabine is higher in 3R3R and similar for 2R allele carriers compared to the FDA approved dose. Objectives 1) To determine the maximal tolerated dose (MTD) and RP2D of capecitabine in EA patients with advanced stage malignancy 2) To determine the safety and toxicity of this regimen and 3) To perform plasma pharmacokinetics (PK) of capecitabine and its metabolites. Methods: EA patients with advanced stage cancer were prospectively allocated into two cohorts: Group A (3R3R) and Group B (2R3R, 2R2R). In each cohort, dose escalation followed a standard phase I 3+3 design. Additional patients were treated at the R2P2 dose level (DL). Initial dose was 1250/m2 bd for 14 days q3w (DL 1) with 125 mg/m2 bd increments subsequently. Pharmacokinetics (PK) of capecitabine and its metabolites were performed using a LC-MS/MS method. Results: 23 patients (Group A=18; group B=5) received 94 cycles (median 2.5, range 1-8). Median age was 58 (range 34-74) years. Median turnaround time for TYMS genotyping was 1 day. In group A, grade 3-4 DLTs were seen in 3 patients: diarhoea, neutropenic fever, hand-foot syndrome, and mucositis. MTD was at DL 4 (1625mg/m2 bd) and the RP2D was 1500mg/m2 bd (DL 3). DL 3 was expanded to n=9. At DL 3, day 1 mean (± SD) capecitabine and 5FU Cmax was 12.3 ± 9.9 and 0.91 ± 0.73 µg/mL, respectively and AUC0 – t was 9.84 ± 5.53 and 1.21 ± 0.54 h*µg/mL, respectively. Compared with DL1, Cmax for capecitabine and 5FU was 2 and 2.02 fold higher and AUC 0-t was 1.49 and 1.59 fold higher at DL3. Accrual in group B was ceased at DL2 due to lack of patient enrolment; no DLT was seen. By ROC analysis, day 1 5FU AUC0 – t of 1.74 h*µg/mL predicted for DLT (p=0.022, sensitivity 100%, specificity 90%). Conclusions: In EA patients with TSER 3R3R, the RP2D was 1500 mg/m2 bd. The D1 5FU AUC0 – t at RP2D was 59% more than the FDA approved dose (DL1).

4100^ Background: Preclinical studies have shown that pharmacological inhibition of the MEK/ERK pathway by AZD6244 enhanced the anti-tumor effect of sorafenib in both orthotopic and ectopic models of HCC. We conducted this study to determine tolerability, pharmacokinetics, and pharmacodynamics of AZD6244 when combined with sorafenib in advanced HCC. Methods: Patients with biopsy-proven unresectable BCLC B/C hepatocellular carcinoma were recruited. Only those with Childs-Pugh A or B (7) liver cirrhosis and without prior systemic therapy were included. Sorafenib at 400mg bd was given 1 week before initiation of AZD6244 which was escalated in subsequent cohorts from 75mg om based on 3+3 design. PK and PD studies and QOL assessments were performed. DCE-MRI imaging was performed to assess tumor vascularity in response to treatment. Results: Fourteen patients were recruited (including 2 replaced). 11 had evaluable disease. Characteristics: all male, all Chinese, 12 were BCLC C stage. Two DLTs were seen out of 6 patients at dose level 1 (AZD6244 at 50mg bd) which were grade 3 fatigue and grade 3 abdominal pain with elevated transaminases. When an additional dose level 1A was added (ADZ6244 at 100mg om), 2 out of 3 patients had DLTs of grade 3 raised aspartate transaminase and grade 3 diarrhea. Thus, the MTD was determined to be AZD6244 at 75mg om when combined with sorafenib 400mg bd. Common toxicities were diarrhea (83%), rash (58%), fatigue (50%), hypertension (42%), anorexia/vomiting/thrombocytopenia (25%). Two patients had reversible LVEF dysfunction and there were no eye toxicities. PK of AZD6244 showed oral clearance of 11.2 + 6.8 L/h and terminal half-life of 6.0 +2.0 h.Objective responses were 3 PR, 6 SD and 2 PD.DCE-MRI measurements demonstrated significant reductions in permeability surface area product (PS, ml/100ml/min) and fractional intravascular blood volume (v1, ml/100ml) seven days after starting sorafenib. No additional antivascular activity was observed when AZD6244 was added to sorafenib. Conclusions: The recommended phase II dose for AZD6244 is 75mg om when combined with sorafenib 400mg bd for advanced HCC patients. This combination is feasible, shows activity and warrants further investigation.

e13002 Background: CPT-11 has complex PK and PD; genotyping of UGT1A1 is recommended for higher doses in *28 carriers (Caucasian) and *6 carriers (Asian) because of a higher risk of toxicity. Raltegravir is mainly metabolized to raltegravir glucuronide by UGT1A1. CPT-11 is mainly eliminated by cytochrome P450 isoform 3A (CYP3A)-mediated oxidation and by esterase cleavage to form SN-38, which is further conjugated by uridine-diphosphate glucuronosyltransferase (UGT) isozymes to the inactive SN-38 glucuronide. We prospectively explored the correlation of midazolam and raltegravir PK with CPT-11 clearance (CL) and SN-38 levels respectively and toxicity in patients undergoing chemotherapy with CPT-11. Methods: Twenty-five Asian patients with advanced cancer received CPT-11 as a 2-weekly 90-minute infusion (180 mg/m2) as part of the FOLFIRI regimen. Subjects were administered raltegravir 400 mg orally (as a UGT1A1 probe) and intravenous midazolam 1 mg (as a CYP3A4 probe) one day before the first dose of their chemotherapy. Blood sampling for drug concentrations was performed for 2 days after probe drug administration. Non-compartmental PK analyses were performed using WinNonLin. Genomic DNA was isolated and screened for the known genetic variants in UGT1A1*6. Results: One patient developed grade 3 diarrhea and 16 developed grade 3-4 neutropenia. The mean CPT-11 CL was 21.9 + 7.8L/h. SN-38G/SN-38 AUC ratio was highly correlated with Raltegravir glucuronide/ Raltegravir AUC ratio (r=0.784 p10Nadir Absolute Neutrophil Count (ANC) (r=-0.397 p10 Nadir ANC. Of the 24 patients with genotyping of UGT1A1*6 done, 17 were wild type (GG), 7 heterozygotes (AG) and 0 were mutant (AA). No differences in toxicity or SN38 pharmacokinetics were observed in the UGT1A1*6 variants. Conclusions: Raltegravir glucuronide ratio is a good predictor of SN-38 glucuronidation, and has potential to predict neutropenia from FOLFIRI treatment. Prospective studies with dose adjustments should be done to develop raltegravir as a probe to optimize CPT-11 therapy.

3002 Background: STAT3 is constitutively activated by growth signaling pathways in many malignancies; in many cell lines inhibition of STAT3 leads to cell death. OPB51602 is a small molecule inhibitor of STAT3 phosphorylation (Tyr705) and activation. Methods: A phase I study of OPB51602 administered for 2 weeks every 3 weeks was initiated to determine the maximum tolerable dose (MTD), evaluate pharmacokinetics (PK), and assess pharmacodynamics effects on STAT3 in peripheral blood mononuclear cells (PBMCs). The starting dose was 2mg/day, and dose escalations to 5mg/d, 10mg/d, 20mg/d, were planned. Single dose PK was done on the first day of administration for 4 days. Dose escalation was based on the “3+3” design, MTD was defined as the dose with at least 2/6 dose limiting toxicities (DLTs) in the first cycle and toxicities were graded by NCICTCv4.0. Results: 32 patients (pt) were treated at doses of 2mg/d (n=7), 4mg/d (n=18), 5mg/d (n=7). The main toxicities observed included nausea/vomiting, diarrhea, peripheral neuropathy and fatigue. 5 mg/d was the MTD, where cycle 1 DLTs of grade 3 diarrhea/dehydration and hyponatremia occurred in 1 patient respectively. One pt developed grade 3 peripheral neuropathy at 4mg/d cohort. PK showed maximal drug levels 2-3 hours after administration, bi-exponential decay, with mean oral clearance of 316.5 +/- 638.9 L/h and long terminal mean half-life of 61.8 +/- 15.9 h on day 17. STAT3 phosphorylation in PBMCs assessed in 6 pts receiving 4mg/d was reduced from 67.4 +/- 17.4% at baseline to 53.0 +/- 18.1% (p=0.001) after administration on day 1. Interestingly, reduction in tumor metabolism by PET CT on day 15 was observed in 4/8 pts receiving 4mg/d. A pt with heavily pretreated adenocarcinoma of lung at 5mg/d dose had partial response; another pt with metastatic endometrial cancer at 4mg/d dose experienced stable disease for 6 cycles. Conclusions: OPB51602 has an MTD of 5mg/d on this schedule, demonstrates inhibition of STAT3 phosphorylation, and evidence of clinical activity. Further proof of concept studies of OPB51602 are warranted.

89 Background: Perioperative chemotherapy has been shown to improve outcomes among patients with locally advanced gastric adenocarcinoma. As docetaxel (D), cisplatin (C), and capecitabine (X) are active agents in gastric adenocarcinoma, we conducted a phase II trial to assess the efficacy and tolerability of preoperative DCX in patients with gastric adenocarcinoma. Methods: We enrolled patients with locally advanced gastric adenocarcinoma (clinical T3/4 or N positive) to preoperative DCX repeated every 21 days for three cycles, followed by surgery. The first cohort of 10 patients received docetaxel 35 mg/m2, cisplatin 35mg/m2 on days 1 and 8, with capecitabine 750 mg/m2 twice daily from day 1 to day 14. A subsequent cohort of another 11 patients had dose modifications for docetaxel and cisplatin, both to 30mg/ m2 on days 1 and 8. Results: Twenty-one patients were recruited. Median age was 61 years. Seventy-one percent of patients completed a total of 3 cycles of planned chemotherapy. Fourteen patients have successfully undergone surgery. R0 resection rate was 88%. Pathological complete response (pCR) and near complete response (pnCR) were both 17%. Preoperative radiological assessment showed a partial response rate of 65% after 2 cycles of treatment. Good pathological response (pCR and pnCR) was associated with better event-free survival (26 months vs 12 months, p=0.031). Common grade 3/4 toxicities were diarrhea (38%), neutropenia (30%), stomatitis (14%) and hypokalemia (14%). However, the diarrhea rate was reduced to 15% with dose modifications as mentioned above after 50% of the first cohort of 10 patients developed grade 3/4 diarrhea during the first cycle of chemotherapy. Febrile neutropenia rate was 14%. There was one treatment related death before surgery but no postoperative mortality. Conclusions: Preoperative DCX is highly active with modest toxicity in locally advanced gastric adenocarcinoma. DCX represents an attractive alternative regimen without the need for protracted infusional drug in perioperative chemotherapy for patients with locally advanced gastric adenocarcinoma. Good pathological response is associated with better outcome.

e14504 Background: Linifanib (ABT-869) is an orally bioavailable, potent, VEGF/PDGF RTK inhibitor. Activity was observed in a phase II trial of pts with HCC (ASCO 2010, abstr 4038). An exploratory ...

4038 Background: Linifanib (ABT-869), a novel potent and selective inhibitor of the VEGF and PDGF platelet-derived growth factor families of RTKs, is designed to inhibit angiogenesis, tumor growth,...

4581 Background: ABT-869 is a novel orally active, potent and selective inhibitor of the vascular endothelial growth factor and platelet derived growth factor families of receptor tyrosine kinases. Results of an interim analysis of a phase 2 trial of ABT-869 in HCC are presented. Methods: An open-label, multicenter phase II trial (M06–879) of oral ABT-869 at 0.25 mg/kg daily in Child-Pugh A (C-PA) or QOD in Child-Pugh B (C-PB) patients (pts) until progressive disease (PD) or intolerable toxicity, is ongoing. Key eligibility criteria included unresectable or metastatic HCC; up to one prior line of systemic treatment; and at least one measurable lesion by computed tomography (CT) scan. The primary endpoint was the progression free (PF) rate at 16 weeks. Secondary endpoints included objective response rate (ORR), time to progression (TTP), progression free survival (PFS) and overall survival (OS). CT scans were assessed centrally and by the investigators; presented results are from central assessment. Results: 44 pts were enrolled from 09/07 to 08/08 at 6 centers internationally, with interim data available for 34 pts. There were 28 C-PA pts (median age, 63.5 y [range, 20- 81]) and 6 C-PB pts (median age, 64.5 y [range, 36–69]) and 73.5% received no prior systemic therapy. For the 19 evaluable C-PA pts included in the per-protocol interim analysis, 8 (42.1%) were progression free at 16 weeks [95% CI 20.3, 66.5]. The estimated ORR was 8.7% [95% CI, 1.1, 28] for the 23 C-PA pts and 0% for the 2 C-PB pts who had at least one post-baseline CT scan reviewed by central imaging. For all 34 pts, median TTP was 112 d [95% CI, 110, -], median PFS was 112 d [95% CI, 61, 168] and median OS was 295 d [95% CI, 182, 333]. The most common adverse events (AEs) for all pts were hypertension (41%), fatigue (47%), diarrhea (38%), rash (35%), proteinuria (24%), vomiting (24%), cough (24%) and oedema peripheral (24%). The most common grade 3/4 AEs for all pts were hypertension (20.6%) and fatigue (11.8%). Most AEs were mild/moderate and reversible with interruption/dose reductions/or discontinuation of ABT-869. Conclusions: ABT-869 appears to benefit HCC patients, with an estimated TTP of 112 days and an acceptable safety profile. Updated results from this ongoing study will be presented. [Table: see text]