Search

Background Cancer of unknown primary (CUP) presenting as bone‐predominant (BCUP) or lymph node‐only disease (LNCUP) represents two clinically distinct subsets of nonvisceral CUP. These present a diagnostic challenge with a large differential of putative primary cancers and defy the “one‐treatment‐fits‐all” approach. Materials and Methods We identified patients with BCUP (n = 29) and LNCUP (n = 63) using a prospectively collected CUP database and tumor registry of patients seen at MD Anderson Cancer Center between 2001 to 2017. Clinicopathological characteristics, treatments, and outcomes were abstracted. A control group of non‐BCUP/LNCUP cases (n = 443) from the database was used for comparison. Kaplan‐Meier method was used to estimate overall survival and compared using log‐rank test. Results In this cohort, 64% and 60% patients had disseminated disease at diagnosis and 39% and 23% had Culine poor‐risk disease in BCUP and LNCUP, respectively. Median overall survival (OS) for BCUP was 14.5 months and for LNCUP was 32.6 months. For BCUP, gemcitabine plus platinum was the most common initial chemotherapy (54%). For LNCUP, carboplatin plus paclitaxel was the most common initial chemotherapy (38%). Radiation was given to 74% of patients with BCUP and 37% of those with LNCUP. On multivariate analysis, poor‐risk Culine group (hazard ratio [HR], 1.76; p < .001) and high neutrophil‐to‐lymphocyte ratio (HR, 2.38, p < .001) were associated with worse OS. Conclusion BCUP and LNCUP are rare subsets within CUP with varying prognosis. Poor‐risk Culine group and high neutrophil‐to‐lymphocyte ratio are associated with poor survival. Select patients with limited metastases can have long‐term survival with aggressive multimodality treatment. Careful clinicopathological review can facilitate chances of site‐directed therapy. Implications for Practice Cancer of unknown primary (CUP) rarely presents as bone‐predominant (BCUP) or lymph node‐only (LNCUP) disease. This article describes a cohort of each and compares with a larger CUP cohort. Patients with BCUP have unique issues with fractures and pain, often receiving radiation. Overall survival of 14.5 months was similar to a larger CUP comparison cohort. Patients with LNCUP had improved overall survival at 32.6 months, with longer survival in patients without disseminated disease. Culine poor‐risk group and neutrophil‐to‐lymphocyte ratio were associated with worse overall survival. Tips regarding diagnosis and management of these rare malignant subsets are provided., Considering the limited understanding of bone‐predominant or lymph‐node‐only cancers of unknown primary, this article describes the clinical features and outcome data of these two unique conditions.

Purpose: Up to 17% of patients with pancreatic ductal adenocarcinoma (PDAC) harbor pathogenic (germline or somatic) mutations in a homologous recombination, DNA damage response and repair (HR-DDR) gene, such as BRCA1/2, or PALB2. Platinum-based chemotherapy, or treatment with PARP inhibitors are of particular benefit in these patients. However, there may be even greater benefit when platinums and PARP inhibitors are combined. Patients and Methods: We performed a single-arm, open-label, phase I/II study of the PARP inhibitor, veliparib, with 5-fluorouracil (no 5FU bolus) and oxaliplatin (FOLFOX) for patients with metastatic PDAC. Thirty-one patients were enrolled in a phase I dose escalation of veliparib (40 mg to 250 mg twice a day, days 1–7 of each 14-day cycle), to identify the recommended phase II dose (RP2D) of veliparib for the combination. Another 33 patients were enrolled in two parallel phase II trials to assess the objective response rate (ORR) in untreated or in previously treated patients. If available, germline or somatic testing was collected to identify pathogenic HR-DDR mutations. Results: The combination of veliparib and FOLFOX was tolerable at a RP2D of veliparib of 200 mg twice a day. The primary endpoint for both phase II cohorts was met, and the ORR overall was 26%. There was greater activity in platinum-naïve patients, and those who harbored a pathogenic HR-DDR mutation. Specifically, the ORR of HR-DDR mutated, platinum-naïve patients was 57%. Conclusions: The combination of veliparib and FOLFOX was safe for patients with metastatic PDAC and showed promising activity particularly in patients with platinum-naïve disease that harbors a pathogenic HR-DDR mutation.

PURPOSE: This phase I trial evaluated the maximum tolerated dose, safety and preliminary efficacy of lapatinib, a HER1, HER2 dual kinase inhibitor plus bortezomib, a proteasome inhibitor, in adult patients with advanced malignancies. METHODS: Patients were enrolled in a standard 3+3 design with lapatinib (L) 750, 1000, 1250 or 1500mg daily, and bortezomib (B) 0.7, 1.0, 1.3 or 1.6mg/m(2) 3 weeks on 1 week off. Dose limiting toxicities (DLT) were assessed during the first 28 days. RESULTS: Fifteen patients received the combination of lapatinib and bortezomib in 3 different cohorts and 10 were evaluable for DLT. There were no DLTs. Anorexia was the most common adverse event. Biomarker analysis showed upregulation of p27 expression with lapatinib and the combination. No tumor response was observed and thus the study was closed early. CONCLUSION: The combination of lapatinib and bortezomib was well tolerated but no complete or partial tumor responses were observed at the dose levels tested. ClinicalTrials.gov Identifier .

Multimodality therapy (MMT) is recommended for patients with resectable gastric cancer, but no single approach has been established as standard. Little is presently known about current national practice patterns and sequencing of MMT.Retrospective cohort study of patients with gastric cancer aged 18 to 80 y in the National Cancer Database (2006-2014) with ≥T2 and/or node-positive disease (i.e., stage Ib to III) treated with MMT. Clinical nodal staging accuracy was ascertained among those treated with upfront surgery by comparing clinical and pathologic nodal staging. Multivariable Cox regression was used to evaluate the association between overall risk of death and MMT approach (i.e., radiation used versus not and treatment sequence).Among 5817 patients, 16.1% received perioperative MMT, 50.6% preoperative only, and 33.3% postoperative only. The sensitivity, specificity, positive predictive value, and negative predictive values of clinical nodal staging were 68.4%, 88.8%, 91.1%, and 62.7%, respectively. Current clinical nodal staging modalities understage 37.3% of clinically node-negative patients. Over time, radiation utilization decreased (74.3% in 2006 versus 53.9% in 2014; trend test, P 0.001), perioperative MMT increased (8.9% versus 22.2%%; trend test, P 0.001), and postoperative MMT decreased (43.1% versus 21.0%; trend test, P 0.001). Neither type of MMT nor treatment sequence is associated with risk of death.One-third of patients with gastric cancer who are candidates to receive MMT are treated with upfront surgery. Given the high false negative rate of clinical nodal staging and high noncompletion rate of postoperative treatment, efforts should be directed at improving and optimizing preoperative therapy utilization.

Purpose: Prognostic uncertainty is a major challenge for cancer of unknown primary (CUP). Current models limit a meaningful patient-provider dialogue. We aimed to establish a nomogram for predicting overall survival (OS) in CUP based on robust clinicopathologic prognostic factors. Experimental Design: We evaluated 521 patients with CUP at MD Anderson Cancer Center (MDACC; Houston, TX; 2012–2016). Baseline variables were analyzed using Cox regression and nomogram developed using significant predictors. Predictive accuracy and discriminatory performance were assessed by calibration curves, concordance probability estimate (CPE ± SE), and concordance statistic (C-index). The model was subjected to bootstrapping and multi-institutional external validations using two independent CUP cohorts: V1 [MDACC (2017), N = 103] and V2 (BC Cancer, Vancouver, Canada and Sarah Cannon Cancer Center/Tennessee Oncology, Nashville, TN; N = 302). Results: Baseline characteristics of entire cohort (N = 926) included: median age (63 years), women (51%), Eastern Cooperative Oncology Group performance status (ECOG PS) 0–1 (64%), adenocarcinomas (52%), ≥3 sites of metastases (30%), and median follow-up duration and OS of 40.1 and 14.7 months, respectively. Five independent prognostic factors were identified: gender, ECOG PS, histology, number of metastatic sites, and neutrophil-lymphocyte ratio. The resulting model predicted OS with CPE of 0.69 [SE: ± 0.01; C-index: 0.71 (95% confidence interval: 0.68–0.74)] outperforming Culine/Seve prognostic models (CPE: 0.59 ± 0.01). CPE for external validation cohorts V1 and V2 were 0.67 (± 0.02) and 0.70 (± 0.01), respectively. Calibration curves for 1-year OS showed strong agreement between nomogram prediction and actual observations in all cohorts. Conclusions: Our user-friendly CUP nomogram integrating commonly available baseline factors provides robust personalized prognostication which can aid clinical decision making and selection/stratification for clinical trials.

The artemisinin class of anti-malarial drugs has shown significant anti-cancer activity in pre-clinical models. Proposed anti-cancer mechanisms include DNA damage, inhibition of angiogenesis, TRAIL-mediated apoptosis, and inhibition of signaling pathways. We performed a phase I study to determine the maximum tolerated dose (MTD) and dose-limiting toxicities (DLTs) of intravenous artesunate (IV AS).Patients were enrolled in an accelerated titration dose escalation study with planned dose levels of 8, 12, 18, 25, 34 and 45 mg/kg given on days 1 and 8 of a 21-day cycle. Toxicities were assessed using the NCI CTCAE (ver. 4.0), and response was assessed using RECIST criteria (version 1.1). Pharmacokinetic (PK) studies were performed during cycle 1.A total of 19 pts were enrolled, 18 of whom were evaluable for toxicity and 15 were evaluable for efficacy. DLTs were seen at dosages of 12 (1 of 6 patients), 18 (1 of 6) and 25 mg/kg (2 of 2), and were neutropenic fever (Gr 4), hypersensitivity reaction (Gr 3), liver function test abnormalities (Gr 3/4) along with neutropenic fever, and nausea/vomiting (Gr 3) despite supportive care. The MTD was determined to be 18 mg/kg. No responses were observed, while four patients had stable disease, including three with prolonged stable disease for 8, 10, and 11 cycles, for a disease control rate of 27%. PK parameters of AS and its active metabolite, dihydroartemisinin (DHA), correlated with dose.The MTD of intravenous artesunate is 18 mg/kg on this schedule. Treatment was well tolerated. Modest clinical activity was seen in this pre-treated population. CLINICALTRIALS.NCT02353026.

Resection techniques for esophageal carcinoma continue to evolve, from endoscopic mucosal resection or endoscopic submucosal dissection for early stage disease to standard and robot-assisted minimally invasive esophagectomy as part of multimodal therapy for locally advanced disease. Though currently limited to assessing conduit perfusion and sentinel lymph nodes, embedded technology in the robotic surgical platform will likely play an expanded role during esophagectomy in the future. The use of targeted therapies, checkpoint inhibitors, engineered immune cell therapy, and cancer vaccines show promise in the treatment of systemic disease. Radiation therapy techniques are becoming increasingly sophisticated and they may play a more active role in stage IV disease in the future.

Purpose To determine the antitumor efficacy and tolerability of combination temozolomide (TMZ) and veliparib (ABT-888) in patients with advanced, sorafenib-refractory hepatocellular carcinoma (HCC). Methods This single-arm phase II trial enrolled patients with pathologically confirmed, sorafenib-refractory HCC. All patients received 40 mg ABT-888 PO daily on days 1–7 and 150 mg/m2 TMZ PO daily on days 1–5 of a 28-day cycle. The primary endpoint was objective response rate (ORR) at 2 months. Secondary endpoints included overall survival (OS), progression-free survival (PFS), and toxicity profile. Tumor response was assessed every 2 cycles using RECIST criteria, and toxicities were assessed using CTCAE v4.03. Results We enrolled 16 patients in the first phase of the trial, but the study was discontinued due to a poor ORR; only four patients (25 %) had SD after 2 cycles. Twelve patients (75 %) were taken off study after 2 months of treatment; 10 of these had disease progression. Two patients (13 %) were taken off study due to severe toxicity, and one patient (6 %) died from non-treatment-related liver failure. One patient had SD for 16 months, receiving 11 cycles of therapy before being taken off study. The most common grade 3 treatment-related toxicities included vomiting (n = 2), thrombocytopenia (n = 2), nausea (n = 1), and anemia (n = 1). The median PFS was 1.9 months, and median OS was 13.1 months. Conclusion The combination of TMZ and ABT-888 is well tolerated in patients with advanced HCC. However, the regimen failed to show survival benefit. ClinicalTrials.gov Identifier NCT01205828. Electronic supplementary material The online version of this article (doi:10.1007/s00280-015-2852-2) contains supplementary material, which is available to authorized users.

BACKGROUND For locally advanced rectal cancer, response to neoadjuvant radiation has been associated with improved outcomes but has not been well characterized in general practice. The goals of this study were to describe disease response rates after neoadjuvant treatment and to evaluate the association between disease response and survival. MATERIALS AND METHODS Retrospective cohort study of patients aged 18-80 y with clinical stage II and III rectal adenocarcinoma in the National Cancer Database (2006-2012). All patients underwent radical resection after neoadjuvant treatment. Treatment responses were defined as follows: no tumor response; intermediate-T and/or N downstaging with residual disease; and complete-ypT0N0. Multivariable, multinomial regression was used to evaluate the association between neoadjuvant radiation use and disease response. Multivariable Cox regression was used to evaluate the association between disease response and overall risk of death. RESULTS Among 12,024 patients, 12% had a complete and 30% an intermediate response. Neoadjuvant chemotherapy alone was less likely to achieve an intermediate (relative risk ratio: 0.70 [0.56-0.88]) or a complete response (relative risk ratio: 0.59 [0.41-0.84]) relative to neoadjuvant radiation. Tumor response was associated with improved 5-y overall survival (complete = 90.2%, intermediate = 82.0%, no response = 70.5%; log-rank, P

Despite its declining incidence, gastric cancer remains a leading cause of cancer-related deaths worldwide. A multimodal approach to gastric cancer is critical to ensure optimal patient outcomes. Pre-therapy fine resolution contrast-enhanced cross-sectional imaging, endoscopic ultrasound and staging laparoscopy play an important role in patients with newly diagnosed ostensibly operable gastric cancer to avoid unnecessary non-therapeutic laparotomies. Currently, margin negative gastrectomy and adequate lymphadenectomy performed at high volume hospitals remain the backbone of gastric cancer treatment. Importantly, adequate gastric cancer surgery should be integrated in the setting of a multimodal treatment approach. Treatment for advanced gastric cancer continues to expand with the emergence of additional lines of systemic and targeted therapies.

716 Background: Pancreatic ductal adenocarcinoma (PDAC) has been highly resistant to immunotherapeutics to date. LOAd703, an oncolytic adenovirus with transgenes encoding TMZ-CD40L and 4-1BBL, has been shown to lyse tumor cells selectively, induce anti-tumor cytotoxic T-cell responses, reduce myeloid-derived suppressor cell (MDSC) infiltration, and induce tumor regression in preclinical studies. Methods: In this phase I/II trial, patients with unresectable or metastatic PDAC are treated with LOAd703 intratumoral injections and standard nab-paclitaxel/gemcitabine (nab-P/G) chemotherapy. Starting on cycle 1 day 15 of nab-P/G, LOAd703 is injected with image guidance into the primary pancreatic tumor or a metastasis every 2 weeks for 6 injections. In the event of sustained tumor control, subjects are eligible to receive 6 more injections. Three dose levels of LOAd703 are being investigated using a BOIN dose escalation design. Primary endpoints are safety and feasibility. Secondary endpoints include response rate and overall survival. Results: To date, 13 subjects are evaluable for safety and feasibility. Three patients were treated at dose 1 (5x10e10 VP), 4 subjects at dose 2 (1x10e11 VP), and 6 subjects at dose 3 (5x10e11 VP). The most common adverse events (AEs) attributed to LOAd703 have been fever, chills, nausea, and increased transaminases. AEs have been transient and grade 1-2, with the exception of a grade 3 transaminase elevation in 1 subject receiving dose 3 (the only dose-limiting toxicity observed thus far). During protocol treatment, circulating MDSCs decreased in 8/13 subjects while effector memory T-cells increased in 10/13. ELISPOT analyses showed a rise in tumor antigen-specific T-cells in 10/13 subjects. At the lowest dose level, best response was stable disease, and 6/10 patients who received higher LOAd703 doses have had partial responses. Only 1 patient has had progressive disease as best response. Conclusions: Adding LOAd703 to nab-P/G has been safe and feasible. Treatment-emergent immune responses have been demonstrated in most subjects, with a notable proportion having objective anti-tumor responses. Clinical trial information: NCT02705196.

4015 Background: 17 – 25% of mPDACs harbor DNA damage response (DDR) mutations, the presence of which can be predictive of a response to platinum and PARP inhibitor-based therapy. The PARP inhibitor, Vel is a potent sensitizing agent for, and has been safely combined with DNA-damaging chemotherapies. Methods: We initiated a Phase I/II trial of Vel + FOLFOX in pts with mPDAC. Pts received standard mFOLFOX6 except without the 5FU bolus, Q2 weeks. For the Phase I portion, a 3+3 dose escalation of Vel identified a recommended Phase II dose of 200mg orally BID, days 1-7, Q2 weeks. For the Phase II portion, we enrolled two cohorts: 1) Untreated pts; 2) Previously treated pts. Also, for Phase II, pts were pre-selected if they had either a pathogenic germline or somatic DDR mutation (e.g. BRCA1/2, PALB2, ATM), and/or a family history suggestive of a breast or ovarian cancer syndrome (labelled FH+). Objective response rate (ORR) was the primary objective of the Phase II cohorts; key secondary endpoints were median progression-free survival (PFS) and overall survival (OS). Results: Between 01-2011 and 12-2018, 64 pts received treatment, 31 in Phase I, and 15 untreated and 18 previously treated in Phase II. The combination was well tolerated, with the main Grade 3/4 AEs being myelosuppression (16%) and nausea/vomiting (6%). Of the 64 pts, 55% were male; median age was 64; 95% had an ECOG PS of 1; 78% were platinum-naïve; 69% were FH+; and 27% had a known DDR mutation. 57 pts were evaluable for response, and the ORR, PFS, and OS for the different pt subgroups are detailed below. The Phase II cohorts achieved the primary endpoint of an ORR ≥ 25%. Most notably, plat-naïve, FH+, and DDR mutation+ pts had an ORR of 58%. Conclusions: The combination of Vel + FOLFOX is safe, well tolerated, and shows promising efficacy particularly in plat-naïve pts who are FH+ and/or harbor DDR mutations. A randomized trial to assess the contribution of Vel to the regimen is warranted. Clinical trial information: NCT01489865. [Table: see text]

To determine the association between neoadjuvant chemotherapy and chemoradiation therapy on completeness of pathologic response and to assess the impact of primary tumor versus nodal response on survival after esophagectomy.Patients aged 18 to 80 years in the National Cancer Data Base (2006-2016) with clinically staged, locally advanced (cT2-4 or cN+) esophageal adenocarcinoma who underwent an R0 esophagectomy after neoadjuvant chemotherapy or chemoradiation therapy were included. Multivariable Cox proportional hazards regression models were constructed to assess the association between treatment response and survival.Among 2870 patients, there was a significant dose-response association between completeness of response and overall survival: no response (reference), partial response (hazard ratio [HR], 0.81; 95% confidence interval [CI], 0.72-0.91), and complete response (HR, 0.55; 95% CI, 0.47-0.65). Compared with neoadjuvant chemotherapy alone, neoadjuvant chemoradiation was associated with higher pathologic primary tumor (33.9% vs 21.3%; P .001) and pathologic nodal response rates (55.9% vs 32.7%; P .001). Both a primary and nodal response were associated with improved survival. However, among patients with a primary but no nodal response, primary tumor response was not associated with risk of death (HR, 0.88; 95% CI, 0.69-1.11). In contrast, among patients who had a nodal but no primary response, the survival benefit of a nodal response was maintained (HR, 0.66; 95% CI, 0.58-0.76).Pathologic nodal (rather than primary tumor) response to neoadjuvant therapy is associated with improved survival. These data suggest a need to optimize neoadjuvant strategies associated with more complete nodal response rates and to consider more aggressive adjuvant treatment for patients with residual nodal disease after esophagectomy.

Gastroesophageal adenocarcinomas represent one of the top five most common types of cancer worldwide. Despite significant advancement, it is still not known which first-line chemotherapy option is best matched to an individual patient. The vast advances in molecular biology have led to the discovery of many potential predictive biomarkers, such as HER-2 neu, thymidylate synthase (TS), excision repair cross-complementation group 1 (ERCC1), and topoisomerase-1 (TOPO1). These markers could allow us to select treatment based on an individual’s tumor profile, resulting in an improvement of outcome. Our report highlights two patients with metastatic gastric cancer that achieved an exceptional response with traditional therapy and provides insights into the future perspectives of molecular profile-directed chemotherapy.

TPS507 Background: No effective therapy is available for patients (pts) with advanced cholangiocarcinoma (CC) after disease progression on first-line chemotherapy. In Phase I studies, pembrolizumab (pem) showed activity in pts with advanced CC. Interferon alpha-2b was shown to increase tumor immune infiltrates. In order to improve the response rate (RR) of pts with CC to pem, we propose to combine sylatron with pem in an open-label, single-arm, multicenter Phase II study. Methods: To be eligible for study, pts must have CC, manifesting as either intrahepatic, extrahepatic or gallbladder cancer that is unresectable, metastatic, and has either failed to respond to or demonstrated progression on frontline chemotherapy. Forty-four pts will be enrolled on study to receive SC sylatron (200 mg) weekly for 12 weeks, and IV pem (200mg) once every 3 weeks, starting on week 4 (at the same time as the 4th dose of sylatron), until pts experience disease progression or unacceptable toxicity, or withdraw consent. CT scans will be performed every 9 weeks. The primary endpoints are overall RR (ORR) and the safety and tolerability of combined pem and sylatron therapy. Secondary endpoints include OS, PFS, and ORR by irRECIST. An initial 3-week window will exist in which to evaluate the effect of sylatron alone (following prior chemotherapy) on the tumor microenvironment: one biopsy (biopsy 1) will be obtained before the start of sylatron therapy, and another (biopsy 2) after the third but before the fourth injection of sylatron (and the initiation of pem plus sylatron combination therapy). The effect of sylatron on the CC immune microenvironment will be studied by comparing biopsies 1 and 2. Immune microenvironment in pem + sylatron responders will be compared with non-responders. The study will assess whether the addition of sylatron improves the RR of pem from 17% to 35% in pts with advanced CC. Simon’s two-stage optimum design will be used to test the null hypothesis that P ≤ 0.17 vs. the alternative hypothesis that P ≥ 0.350. This study design provides 80% power (one-sided significance level of 0.05). The time-to-event endpoints (PFS, OS) will be estimated using Kaplan-Meier methodology.

4035Background: Patients (pts) with esophageal cancer (EsophCa) have a poor prognosis and limited treatment options. The effect of histological subtype on tumor molecular profile remains unknown. H...

TPS3624Background: Maintenance therapy with a fluoropyrimidine plus bevacizumab is a widely accepted strategy in the treatment of mCRC, having been shown significant improvement in progression-free...

2589Background: We demonstrated in mouse xenografts that CDK4/6 inhibition with palbociclib (P) is synergistic with 5FU in suppressing tumor growth. Based on these preclinical data, we initiated a Phase I trial to evaluate the safety and activity of P plus 5FU. Methods: Eligible patients (pts) had advanced, retinoblastoma (Rb)-positive (as demonstrated by IHC) solid tumors with an adequate performance status and intact organ function. The first stage of the study was a dose escalation to identify the recommended phase II dose (RP2D) of P in this combination (n = 21). The second stage is an ongoing schedule optimization, in which the timing of the P and 5FU are varied to identify the most effective sequence, as demonstrated by Ki67 suppression (n = 8 to date). All pts underwent tumor biopsies prior to treatment (tx), after 7 days of P, and again after the 5FU infusion. Results: Six pts were enrolled at a P dose of 50mg orally daily, Days 1-7, and IV 5FU as a 400mg/m2 bolus Day 8, and 2400mg/m2 continuous i...

461 Background: Hepatocellular carcinoma (HCC) is the sixth most common cancer and the second leading cause of cancer-related death worldwide. Most HCCs develop in severely damaged liver. Methods: The effect of multiple treatment options on liver function (LF) and disease outcome of patients (pts) with HCC (n = 185) were examined retrospectively. Pt tumor burden (using Barcelona clinic liver cancer [BCLC] classification) and LF (Child Pugh [CP]) were assessed at time of diagnosis and then after treatment. Using Kaplan Meier with log rank and T tests, BCLC and CP scores were correlated with overall survival (OS) following individual treatment regimens. Results: We show that better BCLC and LF scores at time of diagnosis predict a better outcome (median OS; p < 0.05). Pts received one or more of the following: no treatment, experimental treatment, TACE, Y90 radioembolization, radiofrequency ablation, resection, radiation, and sorafenib (SFB). Considering all treatment scenarios, LF improved in 9.5%, did not change in 33%, and worsened in 57% of pts. Sixty percent of untreated pts experienced LF decline, compared with 33, 54, 48 and 50% of pts receiving TACE, SFB, TACE/SFB, and Y90/SFB, respectively (no significant differences [NS]); 72% of pts receiving TACE/Y90/SFB (NS); and 85% of pts receiving other Y90/SFB combinations (p = 0.006). In general, pts who saw no change or an improvement in LF from baseline had longer median OS versus pts who had declining LF (p < 0.002). However, pts receiving TACE/Y90/SFB (n = 29) had similar OS to those receiving TACE/SFB (n = 35), despite the toxicity difference (72% [TACE/Y90/SFB] vs. 48% [TACE/SFB] of pts had declining LF). TACE/SFB or TACE/Y90/SFB led to longer median OS than any other treatment group (p = 0.017). Conclusions: The outcome of pts with HCC depends on disease stage and LF. Most pts experience LF decline during treatment. Despite LF decline in 48% of pts receiving a SFB/TACE, these pts experienced longer median OS than pts on any other treatment. Balancing survival benefit with liver toxicities is critical to the successful treatment of pts with advanced HCC. Agents with good antitumor activity and minimal liver toxicity are desperately needed for these pts.

444 Background: Stereotactic body radiation therapy (SBRT) is a novel treatment modality for unresectable, locally advanced pancreatic cancer (LAPC). SBRT delivers high doses of radiation over a shorter duration, with minimal acute toxicities, compared with conventional radiation therapy (CRT) using intensity modulated radiation therapy (IMRT) or 3D conformal techniques (3D-CRT). Few studies have compared outcomes between SBRT and CRT techniques. We conducted a retrospective matched pair analysis of LAPC patients receiving chemotherapy and either SBRT or CRT to assess local control (LC), progression-free survival (PFS), and overall survival (OS). Methods: We retrospectively analyzed 28 patients with LAPC who were treated between August 2006 and August 2014 with 5-FU- or gemcitabine-based chemotherapy combined with CRT (45–56 Gy over 25–28 fractions) or SBRT (25–30 Gy over 5 fractions). Fourteen patients treated with CRT were matched with 14 patients treated with SBRT for age (< 65 vs. ≥ 65 years); nodal staging (N0 or N1); and chemotherapy type (gemcitabine-based or 5-FU-based, administered either concurrently or as induction therapy). Tumor response was assessed using RECIST and all outcomes were calculated from the date of diagnosis. The Log Rank test of Kaplan-Meier curves was used for statistical analyses. Results: Median follow-up time was 11 months (mo). The median duration of LC, PFS, and OS for the entire cohort was 11, 8, and 14 mo, respectively. Median duration of LC was 10 mo for CRT and 11 mo for SBRT (p > 0.05). Median PFS was 8 mo for both groups (p > 0.05). The median OS times were 12 mo for SBRT and 16 mo for 3D-CRT/IMRT (p > 0.05). Conclusions: In this small matched-pair analysis, LAPC patients treated with chemotherapy had similar outcomes following 5 fractions of SBRT as compared with a protracted course of CRT. Given the shorter treatment duration with SBRT, further study is warranted with increased cohort sizes and stratification by other patient and disease characteristics.

TPS773 Background: Treatment with Regorafenib (REGO) has shown significant clinical benefits in metastatic colorectal cancer (mCRC) patients (pts) as demonstrated in the CORRECT and CONCUR trials. Results from both studies suggest that subgroups have differential responses. Further research to identify these subgroups through the identification of molecular biomarkers is needed. Methods: Forty pts with refractory mCRC are being enrolled in this study. The primary objective is to prospectively identify tissue and serum-based biomarkers that can predict response to REGO. Secondary objectives are to determine molecular mechanisms by which REGO controls refractory mCRC, as well as molecular pathways involved in the acquisition of resistance. Tumor and blood samples are obtained prior to and 2 weeks after starting REGO. Blood samples are collected on day 1 of each cycle thereafter. Pts will receive 160 mg REGO daily for 3 weeks of each 4-week cycle until disease progression or unacceptable toxicity. Multi-omic based biomarker discovery approaches will be used to uncover predictive marker candidates with special attention to the tumor microenvironment. Laser capture microdissection will be used on tumor tissue to procure highly enriched populations of pt-matched epithelial and stromal/immune cell infiltrates. Each of these entities will be analyzed for RNA expression changes and protein signaling/drug target activation mapping. Protein signaling analysis will be performed by reverse phase protein array of key REGO-related proteins and phosphoproteins (e.g. phosphoVEGFR, Tie2, phosphoRET), as well as broad-scale mapping of mitogenic, survival, autophagy, inflammatory, motility, and signaling networks. Tumor profiling will include next-generation sequencing for 592 genes with 53 selected gene fusions, and IHC and FISH/CISH for selected biomarkers, including PDL1, HER2, MSI, TS, ERCC1, and TOPO1. Blood samples will undergo protein, miRNA, and mutated DNA analysis, as well as exosomal signature study via a proprietary synthetic polyligand multiplexed aptamer-based assay. Exploratory analysis of biomarkers will be used to determine correlations between the presence of, or change in, biomarker levels and clinical response. Clinical trial information: NCT01949194.