Detection and description of ERBB2 amplification using circulating cell free tumor DNA (ctDNA) genomic analysis in metastatic colorectal cancer (mCRC)

661

Background: ERBB2 amplification (HER2amp) is evolving as a distinct subset of mCRC with therapeutic implications. Although seen in 3-4% of mCRC in tumor tissue, little is known about occurrence of HER2amp in ctDNA. The purpose of this study was to assess the value of ctDNA in detection of HER2amp in blood and to delineate its landscape in mCRC. Methods: We performed a retrospective analysis of mCRC patients (pts) who underwent plasma-derived NGS of ctDNA in a CLIA-certified lab (Guardant Health, Inc.) using a high sensitivity assay that reports copy-number variations using pre-specified algorithms. Pts with no detectable alterations were excluded. The cohort was divided into 2 groups: HER2amp and ERBB2 non-amplified (HER2NA). Descriptive statistics and Fisher’s exact test were used. Results: Between 5/2014 and 5/2016, 1625 mCRC pts had a ctDNA NGS assay and of these 1387 met our inclusion criteria. Among these, HER2amp was seen in 69 pts (4.9%, 95%CI: 3.9 – 6.3). HER2amp were found to be enriched in RAS wild-type (WT) (6.4% v 2.9%, OR 2.3, P = 0.002) and RAS/BRAFV600E WT (6.9% v 2.7%; OR 2.7, P < 0.001). When co-existing with RAS mutations, HER2amps were seen more with sub-clonal ( < 50% relative mutation allele frequency) RAS mutations than clonal RAS mutations (6.6% v 2.1%, OR 3.1, P = 0.03). The most common co-occurring mutations were TP53 (77%), APC (59%), EGFR (25%), PIK3CA (25%) and KRAS (23%). HER2 (13% v 5%, OR 2.6, P = 0.017) and TP53 (77% v 62%, OR 2.1, P = 0.011) mutations co-occurred more frequently with HER2amp than with HER2NA cases. No significant association was seen with PIK3CA, EGFR and APC mutation status. Conclusions: In one of the largest series of CRC pts with ctDNA NGS, ERBB2 amplification were confirmed to be enriched in RAS/BRAF WT mCRC. In these pts, co-existing clonal RAS mutations are uncommon, but concurrent PIK3CA and ERBB2 mutations are present in a sizable minority of cases. These findings may have implications for anticipated innate/acquired resistance mechanisms to HER2-targeting therapies under evaluation for CRC.

Lymphopenia and outcomes of patients with unresectable hepatocellular carcinoma treated with radiotherapy

278

Background: The immune system plays a crucial role in cancer surveillance. Lymphopenia during cancer treatment portends poor prognosis. Here, we aim to quantify lymphocyte count changes during radiotherapy (RT) for patients with hepatocellular carcinoma (HCC) treated with proton or photon RT. Methods: Patients with HCC treated with definitive RT from 2006 to 2016 with ≥3 months follow-up were studied. Patient characteristics, tumor features, treatment specifics were recorded. Serial absolute lymphocyte counts (ALC) were collected and graded according to CTCAE v4.0. Overall survival (OS) and distant metastasis free survival (DMFS) were analysed using the Kaplan-Meier method. To investigate difference in ALC between treatment modalities, we performed a paired cohort analysis (15 proton and 15 photon cases, matched to treatment volume). Results: 110 patients were studied. Median age was 67 years. 70% of the cohort had Child-Pugh A5 liver disease and 54% had venous tumor thrombosis (VTT). 75 (68%) had photon RT and 35 (32%) had proton RT. Median OS was 13 months. Overall, there was a 60% drop in ALC during RT. Splenic volume correlated with Grade 3 (G3) lymphopenia (p < 0.0001). Pre-treatment ALC and splenic volume did not correlate with OS or DMFS. Patients who had G3 lymphopenia during RT had poorer DMFS (14 vs 34 months, p = 0.006) and OS (16 vs 34 months, p = 0.015) than their counterparts. Those who received proton RT had a higher ALC nadir (0.56 vs 0.40 k/ul, p = 0.041), lower rate of ALC drop (0.07 vs 0.05 k/ul/day) and better OS (30 vs 22 months) than those who had photon RT. Paired cohort dosimetric analyses revealed that there was a significantly higher low dose volume in the photon group (V1Gy and V10Gy, p < 0.0001 for both; V5Gy, p = 0.002). Higher volume of low dose bath correlated with lower ALC (r = -0.34, p = 0.06 for V1Gy; r = -0.44, p = 0.01 for V5Gy; r = -0.51, p = 0.004 for V10Gy). Splenic irradiation dose correlated with ALC nadir, and proton cases had significantly less splenic dose (p < 0.0001). Conclusions: G3 lymphopenia during RT portends poorer outcomes for patients with HCC. Protons may mitigate RT-induced lymphopenia. An ongoing phase III trial will compare proton and photon RT for HCC to determine the effect on OS.

Predictors for detecting circulating tumor DNA (ctDNA) in metastatic colorectal cancer (mCRC)

634

Background: Utilization of ctDNA has been rapidly adopted as a predictive diagnostic in advanced NSCLC and indications in GI cancers may be emerging. While tissue-based assays have yields above 90%, there is less known about factors that influence the sensitivity of ctDNA for detecting mutations. Methods: We retrospectively evaluated mCRC patients (pts) who had plasma-derived NGS utilizing a highly-sensitive 68-73-gene assay. Tissue from prior resections or biopsies underwent concurrent 46-gene sequencing. In a case-control design, pts with a known mutation on tissue and radiologic evidence of metastatic disease but no detectable ctDNA mutation were matched 1:3 with randomly selected pts with detectable mutations and compared according to clinical, laboratory, and radiologic characteristics. A binary logistic regression was performed and Kaplan-Meier and Cox-proportional hazards models were used to compare overall (OS) and progression-free (PFS) survivals. Results: Of 427 mCRC pts who underwent ctDNA testing, 416 pts met inclusion criteria. Plasma-derived NGS did not find tumor mutations in 66 cases (15.9%); 198 pts with detectable alterations were selected as controls. After multivariate analysis, the lack of detection of ctDNA was associated with decreasing age (OR 0.94; 95%CI 0.91-0.98; p = 0.004), absence of liver (OR 0.19; 95%CI 0.08-0.45; p < 0.001) and lymph node metastases (OR 0.28; 95%CI 0.12-0.70; p = 0.006). A key determinant was timing of collection relative to disease status: plasma collected after evidence of progression was substantially more likely to have detectable alterations (OR 10.95; 95%CI 4.23-28.33; p < .001); in these pts, the modeled rate of detection was 98%, which increased to > 99% if the pts also had either liver or nodal disease. Pts with no detectable ctDNA had better OS (HR 0.38; 95%CI 0.21-0.69, p = 0.002) and PFS (HR 0.52; 95%CI 0.32-0.85; p = 0.009). Conclusions: When limited to ctDNA collected in newly diagnosed or recently progressing pts, the yield of ctDNA is 98%, which equals or exceeds the yield of tissue based testing. Our findings support the notion that ctDNA testing, when appropriately utilized, can replace tissue based testing and may provide prognostic information.

The rate of novel actionable mutations in standard of care NGS panel testing in gastrointestinal malignancies

640

Background: In advanced gastrointestinal (GI) malignancies, genetic profiling is often performed with the goal of facilitating enrollment of patients into clinical trials. While multigene genetic profiling has become the standard of care in many practices, the data on success rate of identifying actionable genomic alterations remain limited. In this study, we aimed to characterize the rate of actionable mutations using larger ( > 150 genes) and smaller ( < 150 genes) panels across different GI malignancies. Methods: We reviewed all reports of formalin-fixed paraffin-embedded clinical specimens sent for next-generation sequencing (NGS-using assays of at least 45 genes) for patients with advanced GI malignancies between 2012-2017 at MD Anderson Cancer Center. Actionable mutations were defined as those matching or informing the use of targeted therapies available in clinical trials, or FDA-approved. These were determined by a precision oncology support team (pct.mdanderson.org), using available literature and functional genomic screens. Novel actionable mutations were defined as those not used in current testing guidelines for GI malignancies. Results: Out of 11968 detected mutations, 3832(32.0%) were deemed to be actionable mutations and the remainder were either in non-actionable genes, deemed benign, or variants of unknown significance. Therefore, 1987 (65.1%) of assays had actionable mutations. When limited to novel actionable mutations, the rate fell to 21.5% (659/3052). Compared to CRC, other GI malignancies were 1.65 times more likely to have a novel actionable mutation (95% CI 1.35-2.00, p< .001). The use of larger and smaller panels did not differ in detecting novel actionable mutations, but larger panels resulted in a 3.5-fold higher number of mutations not deemed clinically actionable. Conclusions: Despite incorporation of NGS in oncology practice for GI malignancies, the success rate of detecting novel actionable mutations beyond those in the current guidelines remains low. Using assays with larger gene numbers does not seem to improve this detection rate. Future studies are required to evaluate the success rate of clinical interventions when actionable alterations are present.

A phase II trial of gemcitabine (G), cisplatin (C), and nab-paclitaxel (N) in advanced biliary tract cancers (aBTCs): Updated survival analysis

350

Background: BTCs are often diagnosed at an advanced stage and have a poor prognosis. The standard therapy for aBTCs is the combination of GC. However, the median overall survival (mOS) is dismal at 11.7 months (mos) with a median progression free survival (mPFS) of 8 mos. Methods: A single arm, phase II study was conducted at MD Anderson and Mayo Clinic Arizona. Patients (pts) with aBTC were treated at initial dose level of G/C/N (in mg/m2) at 1000/25/125 (n = 30) which was reduced to lower doses due to grade 3/4 hematological (heme) toxicity (tox) - G/C/N: 800/25/100 (n = 30). Cycles were q21 days with restaging q3 cycles until progression. PFS was the primary endpoint (endpt). Using a Bayesian hypothesis test-based design, we assumed mPFS of 8 mos under the null hypothesis (H0), 10 mos under the alternative (H1). Secondary endpts included mOS, RECIST v1.1 response rate (RR), safety and CA19-9 response. Results: 60 pts were enrolled with 51 being response-evaluable having received more than 1 cycle of therapy (age: median 60 yrs [range 31-77], ECOG PS 0/1 (22/38), M/F (33/27), intrahepatic cholangiocarcinoma/extrahepatic/gallbladder (38/9/13). Median follow-up was 14 mos and median number of treatment (trmt) cycles = 5.24. Pts at initial dose level had significant grade 3/4 heme tox: neutropenia, febrile neutropenia, anemia, and thrombocytopenia leading to trmt discontinuation in 6/30 pts. After dose reduction to G/C/N (in mg/m2) at 800/25/100, trmt was better tolerated with only 3 pts experiencing grade 4 heme tox. Non-heme tox were grade 3 in 19 pts: nausea/vomiting, diarrhea, thromboembolic event/CVA, hypokalemia, constipation, cystitis, LFT elevations. The mPFS = 11.4 mos (95% CI: 6.1, 16.1) and mOS = 19.2 (95%CI: 13.6, NA), 1-year survival rate 67.6%. 51 pts evaluable for response: disease control rate (PR+CR+SD)-84.3% and RR-39%. 12 unresectable cases were operated post trmt with 1 pathologic CR. Conclusions: The combination of GCN was well tolerated at adjusted doses and demonstrates encouraging efficacy having met its mPFS endpt and an impressive mOS higher than historical control. These results merit evaluating GC +/-N in a randomized controlled study. Clinical trial information: NCT02392637.

Clinical and molecular characterization of early-onset colorectal cancer patients with inflammatory bowel disease

689

Background: Recent reports suggest colorectal cancer incidence is steadily increasing in adults < 50 years old. This population has a number of unique pre-disposing conditions, such as inflammatory bowel disease (IBD), that may alter the disease biology within these patients. Methods: Using a cohort of 1876 patients with metastatic CRC (mCRC), we identified 622 patients aged < 50 years and reviewed their medical records for a diagnosis of IBD. Clinical, pathologic and overall survival (OS) were compared between patients with and without IBD in this age group. Results: Twenty of 622 (3.2%) early-onset mCRC patients had IBD. These patients were more likely to have mucinous or signet ring histologic features (50.0% vs 15.6%, OR 5.43, 95% CI 2.20-13.41, P < 0.0001) and less likely to have APC mutations (15.0% vs 42.7%, P = 0.019). A trend toward an increased prevalence of right-sided location (40.0% vs 24.1%, P = 0.053) and TP53 mutations (85.0% vs 65.1%, P = 0.091) and decreased prevalence of PIK3CA mutations (0.0% vs 15.5%, P = 0.057) and KRAS mutations (35.0% vs 50.0%, P = 0.11) were noted. There were no differences in prevalence of BRAF V600 (P = 0.99), CTNNB1 (P = 0.99), NRAS (P = 0.99), PTEN (P = 0.43), or SMAD4 (P = 0.75) mutations between patients with IBD and other early-onset mCRC. The presence of IBD in early-onset mCRC patients was associated with a worse OS compared to patients without IBD but who had mCRC (HR 1.82, 95% CI 0.90-3.69, P = 0.026). Conclusions: Early-onset mCRC patients with IBD are significantly more likely to have mucinous or signet ring histologic features and have fewer mutations in canonical CRC genes such as KRAS, PIK3CA, and APC. This suggests that this population likely develops through a unique molecular pathogenesis and they may respond differently to therapy. Further molecular characterization of this population is required.

Association of phosphatidylinositol 3-kinase (PI3K) pathway activation with increased immune checkpoint expression in colorectal cancer (CRC) patients

653

Background: PI3K pathway is a known modulator of anti-tumor immune response and is frequently activated in CRC through genetic alterations such as PTEN loss (PTENloss) and PIK3CA mutations (PIK3CAmut). This study aims to determine the impact of these alterations on immune cell infiltration, priming and activation in early stage CRC patients (pts). Methods: Immune infiltrates and checkpoints were evaluated using quantitative immunohistochemistry (IHC) on primary CRC (N = 59) for both center of tumor (CT) and invasive margin (IM). Pts were evaluated by presence or absence of either PTENloss or PIK3CAmut (collectively termed PI3K pathway alterations). Microsatellite unstable (MSI) and stable (MSS) tumors were analyzed separately. Clinicopathologic data was examined for potential associations with PI3K pathway alterations. Separately, mRNA data (Agilent) was obtained for immune related genes from an internal cohort with PTEN and PIK3CA annotation (N = 73). Results: 59 pts comprised IHC cohort (40 MSS, 19 MSI); 23 pts (39%) had PTENloss or PIK3CAmut. In Agilent cohort, 16 of 73 pts (22%) had PI3K pathway alterations. In MSS CRC, these alterations were more common in CMS1 (p = 0.03), on right side (p = 0.048) and with peritumoral lymphocytes (p = 0.031). MSS pts with PI3K pathway alterations had higher PD1 protein expression (p = 0.04), 2.1 and 2.3 times increased density of CD3+ (p = 0.01) and CD8+ (p = 0.04) cells respectively, and higher Granzyme B protein expression (p = 0.04) in the CT. These pts also had higher PDL1 gene expression (p = 0.046). MSS CRC pts with PIK3CAmut similarly had 2 times more PDL1 protein expression in epithelial cells of the IM (p = 0.01). Alternate checkpoints were also increased in pts with PI3K pathway alterations, including higher protein expression of LAG3 in CT (P = 0.046) and higher gene expression of CTLA4, TIM3, and TIGIT (P < 0.05 for all). Conclusions: PI3K pathway activated MSS CRC is associated with increased immune engagement, but also upregulation of key immune checkpoints in early stage tumors resulting in an ineffective immune response. Combination of PI3K pathway inhibition with immunotherapy merits investigation in this subset of pts.

Interplay of grade and stage on survival outcomes in appendiceal adenocarcinoma: Corroboration of the AJCC (8th edition) Cancer Staging Classification

773

Background: Histological grade is a key prognostic factor in appendiceal adenocarcinomas (AAs) and has been incorporated in classification of stage IV AAs (IVA – Grade 1: Well; IVB – Grade 2/3: Moderately/Poorly differentiated) in AJCC Cancer Staging (8th edition). The purpose of the study was to corroborate the prognostic impact of this staging classifier. Methods: We performed a retrospective review of 98 AA patients (pts) at UTMDACC (2013 - 2017). Kaplan-Meier method was used to estimate median overall survival (mOS), compared with log-rank tests with emphasis on stage and grade. Results: The cohort included 18, 8, 14, 47 and 11 cases of stage II, III, IVA, IVB and IVC AAs. Histologically, 75 were mucinous (21 non mucinous) and 19, 35 and 42 were grade 1, 2 and 3 respectively. Median age of cohort was 54 years. Median follow-up was 46 months (m). The mOS of stage II, III, IVA, IVB and IVC was 136, 106, 140, 48, 23 m, respectively (P < 0.0001). The mOS of grade 1, 2 and 3 was 160, 75 and 33 m, respectively (P < 0.0001). In univariate analysis (restricted to stage IV pts), grade, AJCC stage, and cytoreductive surgery (CRS) were found to be associated with OS. In multivariate analyses (restricted to stage IV pts), only AJCC stage (HR 3.9, 95% CI: 1.5 - 10.6, P = 0.005) and CRS (HR 3.4, 95% CI: 1.4 – 8.3, P = 0.009) were found to be independently associated with poor OS. In subgroup of mucinous AAs, mOS of grade 3 (32 m) was significantly shorter than grade 2 (54 m) (P = 0.02). In non-mucinous AAs, mOS of grade 3 and grade 2 pts was 38 and 29 m, respectively (P = 0.97). Conclusions: AJCC staging classification has a strong prognostic value in AAs. Beyond the well differentiated AAs which are regarded as a distinct entity in stage IV disease, moderate and poor differentiation also has strong and dissimilar prognostic impact which appears to be dependent on mucinous or non-mucinous subtype in stage IV AAs. Further efforts are needed to incorporate the complex interplay of all grades and mucinous characteristics within the staging system for better prognostication and management.

Randomized trial of irinotecan and cetuximab with or without vemurafenib in BRAF-mutant metastatic colorectal cancer (SWOG S1406)

3505

Background: Metastatic colorectal cancer (mCRC) patients (pts) with BRAFV600 mutations have poor outcomes with standard of care chemotherapy and rarely respond to the BRAF inhibitor vemurafenib. In preclinical models, blockade of BRAFV600 by vemurafenib (V) causes feedback upregulation of EGFR, whose signaling activities can be impeded by cetuximab (C) with anti-tumor activity augmented by irinotecan (I). Methods: Pts with BRAFV600 mutated and extended RAS wild-type mCRC were randomized to irinotecan (180 mg/m2 IV every 14 days) and cetuximab (500 mg/m2 IV every 14 days) with or without vemurafenib (960 mg PO twice daily). Eligible pts had ECOG PS ≤1, and had received 1 or 2 prior regimens with no prior anti-EGFR agents. Randomization was stratified for prior irinotecan. Crossover from the control arm (IC) to the experimental arm (VIC) was allowed after documented progression. The primary endpoint was progression-free survival (PFS, investigator assessed), with 90% power to detect a HR of 0.5, with two-sided type 1 error of 5%. Results: 106 pts were enrolled (99 eligible, 49 in the experimental arm) from 12/2014 to 4/2016, with median age 62 years, 59% female, and 39% with prior irinotecan therapy. PFS was improved with the addition of vemurafenib (HR 0.42, 95% confidence interval [CI] 0.26 to 0.66, P < 0.001) with median PFS of 4.4 (95% CI 3.6 – 5.7) mos vs 2.0 (95% CI 1.8 – 2.1) months. Response rate was 16% vs 4% (P = 0.08), with disease control rate of 67% vs 22%. In pts with no prior irinotecan, median PFS was 5.7 (95% CI 3.0-6.1) months in the VIC arm vs 1.9 (95% CI 1.7 – 2.1) months in the IC arm. Grade 3/4 adverse events higher in the VIC arm included neutropenia (28% vs 7%), anemia (13% vs 0%), and nausea (15% vs 0%). There was no increase in skin toxicity or fatigue. 23 pts (46%) in the IC arm crossed over at the time of progression, with median PFS from crossover of 6.0 months (95% CI 3.7 – 7.4). Overall survival (OS) data will be mature for ASCO 2017. Conclusions: These results demonstrate the clinical benefits of the VIC triplet (vemurafenib, cetuximab, and irinotecan) in pts with treatment-refractory BRAFV600 mutated mCRC, and support VIC as a potential new treatment option in this molecular subset. Clinical trial information: NCT02164916.

Circulating tumor DNA (ctDNA) utilizing a high-sensitivity panel to detect minimal residual disease post liver hepatectomy and predict disease recurrence

3522

Background: Preliminary data suggests that ctDNA can serve as a marker of minimal residual disease following colorectal cancer (CRC) tumor resection. Applicability of current ctDNA testing is limited by the requirement of sequencing known individual tumor mutations. We explored the applicability of a multi-gene panel ctDNA detection technology in CRC. Methods: Plasma was prospectively collected from CRC patients (pts) undergoing hepatic resections with curative intent between 1/2013 to 9/2016. In a blinded manner 5ml of preoperative (preop) and immediate post-operative (postop) plasma were tested using a novel 30kb ctDNA digital sequencing panel (Guardant Health) covering SNVs in 21 genes and indels in 9 genes based on the landscape of genomic alterations in ctDNA from over 10,000 advanced cancer pts with a high theoretical sensitivity (96%) for CRC. Median unique molecule coverage for this study is 9000 for cfDNA inputs ranging from 10 – 150 ng (media input preop = 27 ng, median input postop = 49 ng) with 120,000X sequencing depth on an IIlumina HiSeq2500. Results: A total of 54 pts underwent liver metastectomies with curative intent with a median follow-up of 33 months. Preop blood was a median of 49 days from last systemic chemotherapy and 3 days prior to surgery; postop blood was a median of 17 days after resection. Tumor mutations from standard of care hotspot multigene panel testing (at MDACC) were identified in 46 of 54 pts (85%). Preop ctDNA mutation detection rate was 80% (43/54) and 44% (24/54) in postop setting, with postop median allele frequency of 0.16% (range 0.01% to 20%). In pts with a minimum of 1 year follow up, sensitivity of postop ctDNA for residual disease was 58% (95%CI; 41%-74%), and specificity was 100% (66%-100%). In 43 patients who underwent successful resection of all visible disease, postop detection of ctDNA significantly correlated with RFS (P = 0.002, HR 3.1; 95% CI 1.7-9.1) with 2-year RFS of 0% vs. 47%. Recurrence was detected in ctDNA a median of 5.1 months prior to radiographic recurrence. Conclusions: The detection of postop ctDNA using an NGS panel-based approach is feasible and is associated with a very high rate of disease recurrence.

Phase II study of panitumumab in KRAS wild-type metastatic adenocarcinoma of the small bowel or ampulla of vater

e15799

Background: Given the benefit noted with epidermal growth factor receptor (EGFR) antibodies in KRAS wild-type colorectal cancer, we conducted a phase II trial to evaluate the efficacy of panitumumab in metastatic small bowel adenocarcinoma (SBA) and ampullary adenocarcinoma (AAC). Methods: This is a single-center open-label single arm two-stage phase II study. The primary objective was response rate (RR). Secondary objectives included overall progression-free survival (PFS), overall survival (OS) and toxicity. Eligibility included: KRAS exon 2 wild-type, ECOG PS 0-1, metastatic disease, adequate organ function and prior progression or intolerance to fluoropyrimidine with oxaliplatin. Panitumumab was administered at a dose of 6mg/kg IV every 14 days. Next generation sequencing panels were used for genomic analysis. Results: 9 patients [M/F 7/2, median age: 61 yrs (range: 40-74), ECOG PS 0/1: 2/7] were enrolled from September 2013 to October 2015. 1 pt had AAC (pancreaticobiliary subtype) and 8 pts had SBA (duodenal in 3, jejunal/ileal in 5). The most common toxicity was grade 1 acneiform rash in all patients. The most common grade 2/3 toxicities were anemia (33%), fatigue (22%), hypomagnesemia (22%) and skin infection (22%). None of the patients had a response, 2 patients had SD while the remaining 7 patients had PD. At a median follow-up time of 16.6 months, the median PFS and median OS were 2.4 months and 5.6 months, respectively. Extended RAS mutational testing identified 0/9 patients with mutations in extended KRAS or NRAS. Further genomic analysis of genes relevant to anti-EGFR activity (BRAF, PIK3CA and ERBB2) identified 2/9 patients with BRAF G469A mutation, 1/9 patients with PIK3CA H1047R mutation, and 0/9 patients with ERRB2 mutations. Conclusions: Although well tolerated, panitumumab had minimal clinical activity in patients with metastatic SBA and AAC. Given recent findings suggesting that right-sided colon cancers (hindgut derivation) benefit less from anti-EGFR therapy compared to left-sided colon cancers (midgut derivation), we propose that our findings may relate to the primarily foregut (proximal duodenum) and midgut (distal duodenum to ileum) derivation of the small bowel. Clinical trial information: NCT01202409.

Phase II study for the evaluation of efficacy of pembrolizumab (MK-3475) in patients with cancer of unknown primary

TPS3103

Background: Cancer of unknown primary is a biopsy proven malignancy for which an anatomic primary remains unidentified after a focused search. It accounts for 3-4 % of all solid cancers and most investigators limit it to epithelial and undifferentiated cancers. Patients with metastatic melanoma and sarcoma are excluded. Sophisticated imaging, robust pathologic evaluation including immunostains, and genomic and proteomic characterization of these cancers have challenged the management of CUP. The paradigm has shifted from empiric platinum based combination doublets to a personalized approach. Nevertheless, without an anatomic primary, clinical trial opportunities are limited. There remains an unmet research need to evaluate the role of immunotherapy, specifically checkpoint blockade drugs in specific subsets of CUP patients. Methods: Adult Patients ≥ 18 years of age with ECOG PS 0-1, must meet the definition of a CUP cancer. Patients must be intolerant and/or refractory to at least one line of established therapy known to provide clinical benefit for their condition within the last 6 months (often, a platinum based therapy for carcinomas). Patients must have either measurable (RECIST 1.1) or evaluable disease. Although not limited to subtypes, there is a signficant interest in enrolling patients with isolated disseminated lymphadenopathy, HPV (+) CUP and those who have an IHC profile of those known cancers for which anti-PD therapy has been approved (lung, renal, others) The primary objective of this trial is to evaluate efficacy by evaluation of non-progression rate (NPR) at 27 weeks (9 cycles) as defined as the percentage of CUP patients who are alive and progression-free at 27 weeks (9 cycles) as assessed by RECIST 1.1. Secondary objectives include evaluating safety and tolerability of pembrolizumab (MK-3475); correlating efficacy, non-progression rate (NPR) at 27 weeks (9 cycles), objective response (CR or PR), progression-free survival (PFS), overall survival (OS) and duration of response (DOR) to PD-L1 status; and identifying imaging characteristics associated with immunological changes in tumor following treatment with pembrolizumab. Enrollment is ongoing. Clinical trial information: NCT02721732.

A phase II study of sorafenib and yttrium-90 glass microspheres for advanced hepatocellular carcinoma, BCLC stage C

4083

Background: Combined use of sorafenib and local therapy for treating unresectable hepatocellular carcinoma (HCC) is not well established. Notably, most common cause of death in HCC is liver failure, therefore we tested the promise of controlling the local tumors even in the setting of advanced/metastatic disease to improve survival. Our study aimed to assess the efficacy and safety of combined use of sorafenib and yttrium-90 resin microspheres (Y90 RMS) in unresectable HCC defined as Barcelona Clinic Liver Cancer class C. Methods: Between October 2013 and August 2016 we enrolled 40 advanced stage HCC patients, 38 patients were treated with sorafenib followed (after 4 weeks) with Y90 RMS at MD Anderson Cancer Center. Survival analysis was done to evaluate median overall survival (OS) and progression-free survival (PFS). We used modified Response Evaluation Criteria in Solid Tumors (RECIST) to assess response to treatment and the Common Terminology Criteria for Adverse Events (CTCAE) v4.0 to evaluate the grading of treatment related toxicity. Results: The majority of our patients were males (74%), white (47%), 66% of patients had underlying liver cirrhosis, 26% had vascular invasion, and 26% had extrahepatic disease. The estimated median OS and 95% confidence interval (CI) in months was 18.46 (12.29 – NA) and the estimated PFS was 12.29 months (5.72 – 18.79). Stable disease (SD) was observed in 44.74% of patients, while 28.95% achieved partial response (PR). Grade III-IV adverse events included fatigue (n = 3), hyperbilirubinemia (n = 2), thrombocytopenia (n = 1), proteinuria (n = 1), hyponatremia (n = 1), elevated liver enzymes (n = 4), hypertension (n = 4), diarrhea (n = 1), nausea (n = 1) and vomiting (n = 2). Conclusions: This is the first prospective study to evaluate sorafenib followed by Y90 in HCC. Our study included patients with metastatic HCC and showed that combined use of sorafenib and Y90 was tolerable and was associated with longer OS and PFS compared to previous studies which evaluated sorafenib alone. However, future randomized phase III studies are warranted to assess sorafenib+/-Y90 in metastatc disease setting. Clinical trial information: NCT01900002.

A prospective biomarker study to assess IGF-1 score ability to sub-stratify Child-Turcotte-Pugh classes and predict response to systemic therapy in hepatocellular carcinoma

e15662

Background: Our recent studies showed that lower insulin like growth factors-I (IGF-I) associated with shorter overall survival (OS) in HCC. Furthermore, integrating IGF-I into Child Pugh Score (CTP) (IGF-CTP) led to better prognostic stratification (Kaseb et al., JNCI 2014). Since CTP class A is the standard criterion for active therapy and trials entry, we aimed at assessing the ability of IGF-CTP to predict systemic therapy outcome. Methods: 78 patients were prospectively enrolled and treated with sorafenib. Pre-treatment blood sample were tested for IGF-I and IGF-CTP was calculated after study completion. Survival analysis was done to measure the estimated median OS and progression free survival (PFS), and log rank test was used to compare PFS and OS between subgroups of IGF-CTP score of patients. Results: For CTP A patients, the estimated median OS in months (95% confidence interval, CI) was 9.1m (5.3 – 19.7) and PFS was 5.6m (3.8 – 7.9). Patients who were reclassified as IGF-CTP (B) (OldA/newB = AB) had significantly shorter OS 5.2m (2.8 - NA) and PFS of 4.3m (2.1 – NA), as compared to patients’ who classified as class A by both scoring systems (AA), who had OS of 11.1m (5.7 – 21.3) and PFS of 7.2 m (3.9 – 15.1), P < .001. Interestingly, patients who classified as CTP-B but IGF-CTP-A ( = BA) had significantly longer OS 10.2 (2.89 – NA) and PFS 8.1 (2.9 – NA), as compared to (BB) patients who had OS of 5.8 (3.2 –NA) and PFS of 5.1 (3.19 – NA), P < .001 Conclusions: Our study concluded that IGF-CTP score was more accurate than original CTP score in predicting survival outcomes of systemic therapy in HCC. If validated, this approach may change the standard stratification criteria for active therapy in routine clinical practice and patient selection for clinical trial entry in HCC.

Not all RAS mutations created equal: Functional and clinical characterization of 80 different KRAS and NRAS mutations

3589

Background: Mutations (mts) in RAS predict lack of response to anti-EGFR therapy in colorectal cancer. Outside the “typical RAS” mts ( KRAS/NRAS Codons 12, 13, 59, 61, 117, 146) cited in guidelines and anti-EGFR labeling, clinical impact of other “atypical RAS” mts is uncertain. Methods: Available literature and databases were surveyed to identify 80 KRAS/NRAS mts. We used the NovellusDx Functional Annotation for Cancer Treatment (FACT) to transfect these RAS mts (repeated a mean of 5.5 times/mt) in a cell-based assay that quantifies nuclear ERK localization as a measure of MAPK pathway activation, and normalized to wild type (WT) transfection. In 963 metastatic colorectal cancer patients (pts) with BRAF WT/ KRAS mutant tumors, overall survival (OS) was evaluated by level of RAS signaling activity. Results: Of the surveyed mutations,96% (45/47) of typical mts and 39% (13/33) of atypical mts increased MAPK pathway activation above WT (range: 107%-211% of WT activity). Within the typical RAS mts, mts in NRAS or exon 3, 4 of KRAS had higher activity than mts in exon 2 (codons 12/13) of KRAS, reaffirming the biologic relevance of expanded RAS testing (median activity of 130% vs 178%, P < 0.001). The median activity of atypical RAS mts was lower than typical RAS mts (110% vs 159%, P < 0.001). Several notable exceptions in atypical RAS mts with high activity levels were KRAS V14I, Q22K, D33E, N116S, and F156L (all > 165% of WT activity). Conversely, within the typical RAS mts in the guidelines, KRAS G13C and K117R were not shown to increase activity significantly above WT. Pts with any RAS mt with MAPK activity above the median of typical mts had a worse OS compared to pts below the median in univariate (HR 1.45, 95% CI 1.04-2.32, P = 0.033) and multivariate models (HR 1.96, 95% CI 1.13-3.42, P = 0.017) that controlled for age, gender, sidedness, and synchronous vs metachronous presentation. Conclusions: Functional characterization confirmed activity of RAS mts in the current guidelines, but also suggested that a subset of atypical RAS mutations have similar levels of activation of the MAPK pathway. Within the subset of pts with RAS mts, those mts resulting in high MAPK activity are associated with notably shorter OS.

A phase II trial of gemcitabine (G), cisplatin (C), and nab-paclitaxel (N) in advanced biliary tract cancers (aBTCs)

4018

Background: BTCs are often diagnosed at an advanced stage and have a poor prognosis. The standard therapy for aBTCs is the combination of GC. However, the median overall survival (mOS) is dismal at 11.7 months (mos) with a median progression free survival (mPFS) of 8 mos. Methods: A single arm, phase II study was conducted at MD Anderson and Mayo Clinic Arizona. Patients (pts) with aBTC were treated at initial dose level of G/C/N (in mg/m2) at 1000/25/125 (n = 27) which was reduced to lower doses due to grade 3/4 hematological (heme) toxicity (tox) - G/C/N: 800/25/100 (n = 33). Cycles were q21 days with restaging q3 cycles until progression. PFS was the primary endpoint (endpt). Using a Bayesian hypothesis test-based design, we assumed mPFS of 8 mos under the null hypothesis (H0), 10 mos under the alternative (H1). Secondary endpts included mOS, RECIST v1.1 response rate (RR), safety and CA19-9 response. Results: 60 pts were enrolled with data on 51 available as of the time of this abstract (age: median 60 yrs [range 31-77], ECOG PS 0/1 (17/34), M/F (30/21), intrahepatic cholangiocarcinoma/extrahepatic/gallbladder (32/8/11). Median follow-up was 11.5 mos and median number of treatment (trmt) cycles = 5. Pts at initial dose level had significant grade 3/4 heme tox: neutropenia, febrile neutropenia, anemia, and thrombocytopenia leading to trmt discontinuation in 6/27 pts. After dose reduction to G/C/N (in mg/m2) at 800/25/100, trmt was better tolerated with only 4 pts experiencing grade 4 heme tox. Non-heme tox were grade 3 in 10 pts: nausea/vomiting, diarrhea, thromboembolic event/CVA, hypokalemia, constipation, cystitis, LFT elevations. In the initial 51 pts, mPFS = 11.4 mos (95% CI: 6.1, not reached) and mOS not reached (estimated > 20 mos, 1-year survival rate 66.7%; 95%CI: 65.9-92.2%). 34 pts evaluable for response: disease control rate (PR+CR+SD)-82.3% and RR-32.3%. 3 unresectable cases were operated post trmt with 1 pathologic CR. Conclusions: The combination of GCN was well tolerated at adjusted doses and demonstrates encouraging preliminary efficacy having met its mPFS endpt and a 1-year survival rate higher than historical control. These results merit evaluating GC +/-N in a randomized controlled study. Clinical trial information: NCT02392637.

RAS heterogeneity as a prognostic marker in metastatic colorectal cancer

586

Background: The impact of intratumor heterogeneity on prognosis in metastatic colorectal cancer (mCRC) is unclear, however relative variant allele frequency (rVAF) of key mutations within a tumor may impact outcomes. Therefore, we sought to determine whether rVAF of RAS ( KRAS & NRAS) mutant (mt) clones impacts overall survival (OS) in mCRC patients (pts). Methods: Using a next generation sequencing panel of 201 cancer related genes, we tested 200 mCRC tumors / matched normals. Mutations, indels, and copy number variant (CNV) information were obtained. An rVAF of RAS clones was determined by dividing RAS mt VAF by the VAF of the mutated gene with the highest allele frequency. This truncal gene served as a marker of the total malignant population in a specimen. Pts were stratified at an rVAF of 50%. OS was compared with Kaplan-Meier curves, the log-rank test, and Cox regression. We assessed the impact of CNV on our findings by correcting the rVAF for CNVs in RASand truncal mutations. Results: Of 200 pts, 15% had RAS mt rVAF < 50%, 40.5% had rVAF ≥ 50%, and 44.5% were RAS wild type (WT). Age, gender, MSI status, histology, and stage at diagnosis were similar between groups. More RAS WT pts had BRAF mutations (19.1% vs 1.2% and 3.3%, P< 0.0001), left sided (78.7% vs 56.8% and 60%, P= 0.02), or poorly differentiated tumors (27.3% vs 8.6% and 13.3%, P= 0.003) compared to pts with rVAF ≥ 50% or rVAF < 50%, respectively. Mean coverage was 807x for RAS and 602x for truncal mutations. OS was better in pts with an rVAF < 50% compared to pts with rVAF ≥ 50% regardless of whether rVAF was corrected for CNV (HR 0.6; 95% CI 0.39-0.93, P =0.029) or not (HR 0.48; 95% CI 0.31-0.82, P= 0.010). mOS for pts with WT, rVAF < 50% and rVAF ≥ 50% tumors were 65.8, 55.7, and 38.6 months ( P= 0.0025). In multivariate models controlling for stage at diagnosis and BRAF mutation, pts with rVAF < 50% (HR 1.75; 95% CI 1.03-2.97, P = 0.04) and rVAF ≥ 50% (HR 2.46; 95% CI 1.66-3.65, P< 0.0001) had worse OS compared to WT pts. When rVAF was used as a continuous variable, every 1% increase in rVAF RAS mt resulted in a 1% increased hazard of death ( P <0.0001). Conclusions: Our findings suggest that clonal proportion of a tumor with a RAS mutation may impact OS and suggest the prognostic impact of RAS mutations is not an “all or none” phenomenon.

Signet ring cell colorectal cancer: Genomic insights into a rare subpopulation of colorectal adenocarcinoma

606

Background: Colorectal signet ring cell carcinoma (SRCC) has been shown to be associated with advanced tumor stage at presentation and worse outcomes. Due to the rarity of this subtype, 1% of all colorectal adenocarcinoma (CRC), little is known about its molecular characteristics. We aimed to characterize the molecular alterations of this subgroup. Methods: Metastatic CRC (mCRC) patients (pts) with signet ring cell (SC) histology who had tumors evaluated with next generation sequencing between February 2009 and November 2015 were reviewed. SC mCRC were classified into 2 groups; SRCC (>50% of signet cells) and adenocarcinoma (AC) with SC component. Genomic alterations, microsatellite instability (MSI) and CpG island methylator phenotype (CIMP) status noted in SC mCRC were compared to non-SC mCRC pts from the Assessment of Targeted Therapies Against Colorectal Cancer program at MD Anderson Cancer Center using Pearson’s χ 2 test. Results: A total of 665 mCRC pts were included in this study. 93 pts (14%) had SC histology of which 30 (32.3%) pts were SRCC. The Table below shows key cancer genes mutation frequencies. Conclusions: Colorectal SRCC has distinct molecular features compared with non-SC and AC with SC component CRC. The frequencies of KRAS, PIK3CA and APC mutations were lower than the frequencies reported in non-SC CRC. SRCC was not associated with MSI-H or CIMP-H tumor in this study. Further studies on identification of activated pathways underlying this worse prognosis and potential therapeutic targets are required. [Table: see text]

Cancer of unknown primary (CUP) with disseminated-lymphadenopathy presentation: Clinicopathological features, prognostic factors, and survival outcomes

465

Background: Aside from traditional favorable localized-nodal CUP (LNC) subsets including solitary nodal disease, women with isolated axillary adenocarcinoma and squamous cell carcinoma of cervical/inguinal lymph nodes, there is limited data on disseminated nodal CUP (DNC) presentation. Our study describes the clinicopathologic features and outcomes in nodal CUP (NCUP) highlighting the DNC subset. Methods: We performed a retrospective review of 61 NCUP patients (pts) at UTMDACC (2008 - 2015). For control group (CG), we utilized a cohort of consecutive non-NCUP pts (N = 201) from MDACC CUP database (2012 - 2013). Kaplan-Meier method was used to estimate median overall survival (mOS) and compared with log-rank tests. Results: Median age of NCUP was 61 y, similar to CG (59 y; P = 0.26). In this cohort, 19/61 (31%) and 42 (69%) were classified as LNC and DNC. DNC pts were predominantly women (65%), had good performance status (0/1: 90%), and presented with intrabdominal lymphadenopathy (79%). Sixty percent had carcinoma and majority were high grade (74%). Median immunohistochemical stains performed were 9; most common positive stains were CK7 and pan-keratin. The mOS for entire NCUP cohort was 76 months (m). The mOS for LNC was not reached and was 33.1 m for DNC (HR 0.36; P = 0.04). The mOS of DNC was better than CG (18.8 m; HR 0.62; P = 0.03). Male sex and high neutrophil-lymphocyte ratio (NLR > 5) were poor prognostic factors in DNC ( P < 0.01). Among NCUP pts, 21 reported subsequent development of extra-lymphatic disease [peritoneum/retroperitoneum (29%), liver (24%), lung (19%), muscle/soft tissue (19%), bone (10%)]. Among 36 pts, who had first-line systemic therapy, combination platinum-taxane was used substantially (44%) and was associated with better mOS compared to other regimens (HR 0.20; P < 0.01). Conclusions: Disseminated–lymphadenopathy CUP presentation is a unique clinical subset of CUP which is associated with a significantly better survival than non-NCUP. This distinct entity warrants dedicated research efforts towards understanding biology, developing a cost effective integrated clinical algorithm for management and developing novel therapies.

Association of elevated alpha-1 antitrypsin with advanced clinicopathologic features of hepatocellular carcinoma

289

Background: Alpha-1 Antitrypsin (A1AT) is a circulating liver derived protease inhibitor. There is an evolving evidence that elevated level of A1AT stimulate tumor cell proliferation, and invasion in different cancers. Despite A1AT well-known involvement in hepatic fibrosis, its role in hepatocellular carcinoma (HCC) pathogenesis is not well characterized. The current study aimed to investigate the association between A1AT and clinicopathologic features and prognosis of patients with HCC. Methods: Between 2001 and 2014, total of 766 HCC patients from MD Anderson Cancer Center were enrolled. Under IRB approval, baseline patients’ clinical characteristics were retrieved from medical records. The normal level of plasma A1AT was defined based on the Mayo clinic reference value (1 – 1.9 mg/ml). Survival analysis included Kaplan Meier statistic and Cox regression analysis. Multivariate Hazard Ratio (HR) and 95% Confidence interval (CI) were estimated to determine the independent effect of A1AT on HCC prognosis. Results: The mean and standard deviation of plasma A1AT level was 2.7 ± 0.98 mg/mL. All patients were categorized into 2 groups: group 1 (N = 156) with normal serum level ( ≤ 1.9) and group 2 (N = 610) with higher values ( > 1.9). Median survival (months), 95% CI were 24.4 (18.02 – 30.7) and 11.6 (9.6 – 13.6) in group 1 and 2 respectively, (P < .0001). Patients in group 2 experienced poor clinical characteristics than group 1. The estimated multivariate HR (95% CI) for A1AT is 1.4 (1.1 – 1.7) after adjustment for age, sex, race, cirrhosis, AFP, TNM staging, and treatment exposure. Conclusions: High plasma level of A1AT is associated with higher α-feto protein, advanced TNM and Barcelona clinic liver cancer (BCLC) staging and poor survival of HCC patients. Recent preliminary studies suggested that changes in glycosylaion of production of A1AT by HCC cells correlates with the microenvironment inflammatory and proteolytic activities, which are probably linked to advanced clinicopathologic features and poorer survival. Future excremental studies are warranted to understand the mechanistic pathways of potential A1AT involvement in HCC initiation and progression.

Pertuzumab + trastuzumab for HER2-amplified/overexpressed metastatic colorectal cancer (mCRC): Interim data from MyPathway

676

Background: Among recent advances in precision medicine, HER2 has emerged as a potential therapeutic target for advanced colon cancer. MyPathway is a multi-basket study evaluating the efficacy and safety of targeted therapies in non-indicated tumor types harboring relevant genetic alterations. We present updated data for an expanded cohort of patients with HER2-amplified/overexpressed mCRC receiving HER2-targeted therapy with pertuzumab + trastuzumab. Methods: MyPathway (NCT02091141) is a multicenter, open-label, phase IIA study. Patients in this analysis had treatment-refractory HER2-amplified/overexpressed mCRC by gene sequencing, FISH, or IHC. Patients received standard doses of pertuzumab + trastuzumab until disease progression or unacceptable toxicity. The primary endpoint is investigator-assessed overall response rate. The cutoff date was July 31, 2016. Results: Of 34 patients with mCRC enrolled, 32 have had ≥1 tumor assessment. At a median follow-up of 5.2 (range 0.7–18.3) months from treatment initiation, 12 patients had partial responses (PR), with stable disease (SD) for >4 months in 3 additional patients (Table). The safety profiles were consistent with the product labels. Conclusions: Interim data show that HER2-targeted therapy with pertuzumab + trastuzumab, a chemotherapy-free regimen, is active in heavily pretreated HER2-amplified/overexpressed mCRC. The ORR was 37.5%, responses were durable (median 11.1 months), and the CBR was 46.9%. Accrual to MyPathway is ongoing. Clinical trial information: NCT02091141. [Table: see text]

Phase II trial of bevacizumab and erlotinib as a second-line therapy for advanced hepatocellular carcinoma

Trial registry

Clinicaltrials.gov #NCT01180959.

Background

Early clinical studies of bevacizumab and erlotinib in advanced hepatocellular carcinoma (HCC) have a tolerable toxicity and a promising clinical outcome. We evaluated the efficacy and tolerability of this combination as a second-line therapy for HCC refractory to sorafenib.

Methods

For this single-arm, Phase II study, we recruited patients with Child–Pugh class A or B liver disease, Eastern Cooperative Oncology Group performance status 0–2, and advanced HCC that was not amenable to surgical or regional therapies and treatment with sorafenib had failed (disease progressed or patient could not tolerate sorafenib). Patients received 10 mg/kg intravenous bevacizumab every 14 days and 150 mg oral erlotinib daily for 28-day cycles until progression. Tumor response was evaluated every two cycles using Response Evaluation Criteria in Solid Tumors. The primary end point was the 16-week progression-free survival rate. Secondary end points included time to progression and overall survival.

Results

A total of 44 patients were enrolled and had a median follow-up time of 33.8 months (95% confidence interval [CI]: 23.5 months – not defined). The 16-week progression-free survival rate was 43% (95% CI: 28%–59%), median time to progression was 3.9 months (95% CI: 2.0–8.3 months), and median overall survival duration was 9.9 months (95% CI: 8.3–15.5 months). Grade 3–4 adverse events included fatigue (13%), acne (11%), diarrhea (9%), anemia (7%), and upper gastrointestinal hemorrhage (7%).

Conclusion

Bevacizumab plus erlotinib was tolerable and showed a signal of survival benefit in the second-line setting for patients with advanced HCC. Because standard-of-care options are lacking in this setting, further studies to identify predictors of response to this regimen are warranted.

Retrospective study of non-mucinous appendiceal adenocarcinomas: Role of systemic chemotherapy and cytoreductive surgery

263

Background: The majority of studies evaluating appendiceal epithelial neoplasms have focused on those with mucinous histology. Few studies have reported on non-mucinous appendiceal adenocarcinomas. We performed the largest single-center study to investigate this histologic subtype, in order to describe the natural history and impact of both cytoreductive surgery (CRS) and systemic chemotherapy. Methods: We retrospectively reviewed 172 pts with non-mucinous appendiceal adenocarcinoma evaluated at the UT-MD Anderson Cancer Center between 1990 and 2015. Patient demographics, tumor characteristics, therapy received, and outcomes were recorded. Response assessment was semi-quantitative (response vs. no response) according to the treating physician. Overall survival (OS) and time to progression (TTP) were calculated using Kaplan Meier product-limit method and survival rates compared using the log rank test. Results: Median age at diagnosis was 52.9 yrs (M:F 1:1). Most pts presented with advanced stage: stage I (1.7%), stage II (32.5%), stage III (14.5%), and stage IV (51.2%). No patient had well-differentiated histology. 56% had moderate and 44% poor histology. Median OS by stage was 90.9m [95% CI: 70.8 to 172.9] for stage II, 52.1m [95% CI: 28.9 to NA] for stage III and 28.3m [95% CI: 22.9 to 31.9] for stage IV, (p < 0.0001). In pts with metastatic disease (n = 128) CRS was attempted in 20 (15.6%) and was complete (CCS 0/1) in 12. The median OS for pts achieving complete CRS was 48.6m. Systemic chemotherapy was administered to 92% (118/128) of metastatic pts. The median TTP was 9.4m [95% CI: 8.0 to 11.5] and semi-quantitative response rate was 54%. The majority of pts received either oxaliplatin-based, 57%, or irinotecan-based, 23%, first-line chemotherapy regimens. No statistical difference in TTP (p = 0.9) or OS (p = 0.07) between different chemotherapies was seen. Conclusions: In contrast to mucinous appendiceal neoplasms, non-mucinous appendiceal adenocarcinomas rarely present with low-grade (well-differentiated) histology. Treatment approaches appear more akin to colorectal cancer with most metastatic pts undergoing systemic chemotherapy and a minority undergoing CRS.

Phase II study of bevacizumab combined with capecitabine and oxaliplatin in patients with advanced adenocarcinoma of the small bowel or ampulla of vater

144

Background: Capecitabine with oxaliplatin (CAPOX) is used in front-line management of metastatic small bowel adenocarcinoma (SBA) and ampullary adenocarcinoma (AAC). In a prior single arm phase 2 study (n = 30), CAPOX showed an overall response rate (ORR) of 52%. The purpose of this study was to evaluate safety and activity of bevacizumab and CAPOX in this population. Methods: We conducted a single-center open-label single arm phase 2 study in patients with treatment naïve metastatic SBA and AAC between 8/2011 to 11/2014. Prior fluoropyrimidine-based adjuvant therapy was allowed if completed ≥ 52 weeks. Bevacizumab-CAPOX was administered as a 21-day cycle with capecitabine 750 mg/m2 orally twice daily on days 1-14, oxaliplatin 130mg/m2intravenously on day 1 and bevacizumab 7.5mg/kg intravenously on day 1. The primary endpoint was progression-free survival (PFS) at 6 months. Secondary objectives included ORR, overall PFS, overall survival (OS) and toxicity. Results: A total of 30 patients with median age of 63 years (range 33-78) were enrolled. Six (20%) patients had AAC and 24 (80%) had SBA (duodenal: 18). Treatment was well tolerated and most common grade 3/4 toxicities included fatigue (23%), hypertension (23%), neutropenia (30%), and diarrhea (10%). Common grade 2 toxicities included anorexia (50%), fatigue (47%), and nausea (37%). 4 patients remain on study treatment to date. In 29 evaluable patients the ORR was 48.3% (1 complete and 13 partial responses). Eleven (37.9%) patients had stable disease. The PFS at 6 months was 68% (95% CI: 52%-88%). The median PFS and median OS were 8.7 (95% CI: 6.6-13.4) and 14.98 (95% CI: 10.2-21.3) months, respectively. Comparing the current study to a subset of 25 patients with metastatic disease from our prior phase 2 CAPOX study demonstrated similar ORR, 48.3% vs. 52% (p = 0.79) and PFS 8.7m vs. 6.6m (p = 0.73). Conclusions: In patients with untreated metastatic SBA and AAC, therapy with bevacizumab and CAPOX was well-tolerated and showed efficacy comparable to published regimens. Further studies are warranted to evaluate the benefit of targeting angiogenesis in this rare tumor.

Demographic, tumor characteristics, and outcomes associated with SMAD4 mutation in colorectal cancer

565

Background: SMAD4 regulates signaling in TGF-β pathway as inactivation of SMAD4 renders resistance of cells to TGF-β1-induced growth inhibition. Sporadic mutation of SMAD4 has been reported to be present in 8.6-25% of colorectal cancers (CRC), but the clinicopathologic features and outcomes associated with this mutation have not been described. Methods: Data for patients (pts) with metastatic or locally advanced unresectable CRC who received treatment at MD Anderson Cancer Center whose tumors underwent genotyping for SMAD4 mutation were reviewed and clinicopathologic characteristics, and survival outcomes were evaluated. Tumors were sequenced using a hotspot panel (Ion Torrent, Life Technology) predicted to cover 80% of the reported mutations in SMAD4, and further targeted resequencing that included full-length SMAD4 was performed on mutated ones by HiSeq (Illumina) with full exome coverage to an average depth of 800. Results: Among 616 pts (47.4% females) with CRC, 11.2% of pts (n = 69) had SMAD4 mutation by hotspot testing. SMAD4 mutation was associated with colon cancer versus rectal cancer (OR = 2.1, p = 0.01) and female gender (15% vs 8%, OR = 2.12, p = 0.004). There was no association between the presence of SMAD4 mutation and age, stage at the presentation, tobacco history, colonic location, presence of distant metastasis, histology, nor tumor grade. When compared to pts with wild-type SMAD4, those with SMAD4 mutation had shorter survival from date of testing (median survival of 20.2 months versus 14 months, respectively; log-rank, HR = 1.4, p = 0.014). Full-length sequencing was performed on mutated tumors and it confirmed that missense mutations in R361 and P356 in the MH2 domain were the most common, which are predictive of disruption of both homo- and hetero-oligomerization required for activation. Conclusions: This study is the largest retrospective study to date characterizing SMAD4 mutation in metastatic CRC, and demonstrates a prognostic role for this subgroup of CRC. The prevalence and spectrum of SMAD4 mutations is consistent with previous studies and data from TCGA. Further studies are required to evaluate the implication of the dysregulated TGF- β pathway on response to therapy.

From Protocols to Publications: A Study in Selective Reporting of Outcomes in Randomized Trials in Oncology

PURPOSE

The decision by journals to append protocols to published reports of randomized trials was a landmark event in clinical trial reporting. However, limited information is available on how this initiative effected transparency and selective reporting of clinical trial data.

METHODS

We analyzed 74 oncology-based randomized trials published in Journal of Clinical Oncology, the New England Journal of Medicine, and The Lancet in 2012. To ascertain integrity of reporting, we compared published reports with their respective appended protocols with regard to primary end points, nonprimary end points, unplanned end points, and unplanned analyses.

RESULTS

A total of 86 primary end points were reported in 74 randomized trials; nine trials had greater than one primary end point. Nine trials (12.2%) had some discrepancy between their planned and published primary end points. A total of 579 nonprimary end points (median, seven per trial) were planned, of which 373 (64.4%; median, five per trial) were reported. A significant positive correlation was found between the number of planned and nonreported nonprimary end points (Spearman r = 0.66; P < .001). Twenty-eight studies (37.8%) reported a total of 65 unplanned end points; 52 (80.0%) of which were not identified as unplanned. Thirty-one (41.9%) and 19 (25.7%) of 74 trials reported a total of 52 unplanned analyses involving primary end points and 33 unplanned analyses involving nonprimary end points, respectively. Studies reported positive unplanned end points and unplanned analyses more frequently than negative outcomes in abstracts (unplanned end points odds ratio, 6.8; P = .002; unplanned analyses odd ratio, 8.4; P = .007).

CONCLUSION

Despite public and reviewer access to protocols, selective outcome reporting persists and is a major concern in the reporting of randomized clinical trials. To foster credible evidence-based medicine, additional initiatives are needed to minimize selective reporting.

Obesity Early in Adulthood Increases Risk but Does Not Affect Outcomes of Hepatocellular Carcinoma

BACKGROUND & AIMS

Despite the significant association between obesity and several cancers, it has been difficult to establish an association between obesity and hepatocellular carcinoma (HCC). Patients with HCC often have ascites, making it a challenge to determine body mass index (BMI) accurately, and many factors contribute to the development of HCC. We performed a case-control study to investigate whether obesity early in adulthood affects risk, age of onset, or outcomes of patients with HCC.

METHODS

We interviewed 622 patients newly diagnosed with HCC from January 2004 through December 2013, along with 660 healthy controls (frequency-matched by age and sex) to determine weights, heights, and body sizes (self-reported) at various ages before HCC development or enrollment as controls. Multivariable logistic and Cox regression analyses were performed to determine the independent effects of early obesity on risk for HCC and patient outcomes, respectively. BMI was calculated, and patients with a BMI of 30 kg/m(2) or greater were considered obese.

RESULTS

Obesity in early adulthood (age, mid-20s to mid-40s) is a significant risk factor for HCC. The estimated odds ratios were 2.6 (95% confidence interval [CI], 1.4-4.4), 2.3 (95% CI, 1.2-4.4), and 3.6 (95% CI, 1.5-8.9) for the entire population, for men, and for women, respectively. Each unit increase in BMI at early adulthood was associated with a 3.89-month decrease in age at HCC diagnosis (P < .001). Moreover, there was a synergistic interaction between obesity and hepatitis virus infection. However, we found no effect of obesity on the overall survival of patients with HCC.

CONCLUSIONS

Early adulthood obesity is associated with an increased risk of developing HCC at a young age in the absence of major HCC risk factors, with no effect on outcomes of patients with HCC.

Joint prognostic effect of obesity and chronic systemic inflammation in patients with metastatic colorectal cancer

BACKGROUND

Obesity is strongly linked with chronic systemic inflammation, and each has been linked with disease progression and survival in patients with colorectal cancer (CRC). The authors investigated the joint prognostic effects of obesity and circulating cytokines in patients with metastatic CRC (mCRC), an understudied patient group.

METHODS

In 242 chemotherapy-naive patients with mCRC, the authors measured a multiplex cytokine panel and abstracted clinicopathological features, height, and weight from medical records. Overall survival (OS) was calculated from the date of mCRC diagnosis until the date of death from any cause and evaluated by Kaplan-Meier analysis and multivariable Cox proportional hazards regression models. Cut points for cytokines were determined by restricted cubic spline regression.

RESULTS

In multivariable models, elevated interleukin (IL)-8, IL-2 receptor alpha, and lactate dehydrogenase (LDH) emerged as significant predictors of poor OS (hazard ratio [HR] and 95% confidence interval [95% CI] for above vs below the (referent) knot point: 2.5 [95% CI, 1.7-3.7], 1.9 [95% CI, 1.3-2.7], and 2.2 [95% CI, 1.6-3.1], respectively; all P<.001). Obesity (body mass index ≥30 kg/m(2) ) was not found to be associated with OS, but appeared to modify the relationships observed with IL-8 and LDH, which were associated with a significant 4-fold and 5-fold risk of death, respectively, in obese patients compared with a 2-fold risk of death in nonobese patients (P for interaction of .06 and .04, respectively). Similar results emerged from joint effects analysis, in which obese patients with high IL-8 (or LDH) experienced the highest risk of death.

CONCLUSIONS

Although obesity itself was not found to be independently associated with survival in patients with mCRC, the adverse prognostic significance of LDH and IL-8 was found to be enhanced in obese patients.

Role of HGF/MET axis in resistance of lung cancer to contemporary management

Lung cancer is the number one cause of cancer related mortality with over 1 million cancer deaths worldwide. Numerous therapies have been developed for the treatment of lung cancer including radiation, cytotoxic chemotherapy and targeted therapies. Histology, stage of presentation and molecular aberrations are main determinants of prognosis and treatment strategy. Despite the advances that have been made, overall prognosis for lung cancer patients remains dismal. Chemotherapy and/or targeted therapy yield objective response rates of about 35% to 60% in advanced stage non-small cell lung cancer (NSCLC). Even with good initial responses, median overall survival of is limited to about 12 months. This reflects that current therapies are not universally effective and resistance develops quickly. Multiple mechanisms of resistance have been proposed and the MET/HGF axis is a potential key contributor. The proto-oncogene MET (mesenchymal-epithelial transition factor gene) and its ligand hepatocyte growth factor (HGF) interact and activate downstream signaling via the mitogen-activated protein kinase (ERK/MAPK) pathway and the phosphatidylinositol 3-kinase (PI3K/AKT) pathways that regulate gene expression that promotes carcinogenesis. Aberrant MET/HGF signaling promotes emergence of an oncogenic phenotype by promoting cellular proliferation, survival, migration, invasion and angiogenesis. The MET/HGF axis has been implicated in various tumor types including lung cancers and is associated with adverse clinicopathological profile and poor outcomes. The MET/HGF axis plays a major role in development of radioresistance and chemoresistance to platinums, taxanes, camtothecins and anthracyclines by inhibiting apoptosis via activation of PI3K-AKT pathway. DNA damage from these agents induces MET and/or HGF expression. Another resistance mechanism is inhibition of chemoradiation induced translocation of apoptosis-inducing factor (AIF) thereby preventing apoptosis. Furthermore, this MET/HGF axis interacts with other oncogenic signaling pathways such as the epidermal growth factor receptor (EGFR) pathway and the vascular endothelial growth factor receptor (VEGFR) pathway. This functional cross-talk forms the basis for the role of MET/HGF axis in resistance against anti-EGFR and anti-VEGF targeted therapies. MET and/or HGF overexpression from gene amplification and activation are mechanisms of resistance to cetuximab and EGFR-TKIs. VEGF inhibition promotes hypoxia induced transcriptional activation of MET proto-oncogene that promotes angiogenesis and confers resistance to anti-angiogenic therapy. An extensive understanding of these resistance mechanisms is essential to design combinations with enhanced cytotoxic effects. Lung cancer treatment is challenging. Current therapies have limited efficacy due to primary and acquired resistance. The MET/HGF axis plays a key role in development of this resistance. Combining MET/HGF inhibitors with chemotherapy, radiotherapy and targeted therapy holds promise for improving outcomes.

Multiple independent methods fail to confirm MET amplification rate reported in literature for metastatic colorectal cancer (mCRC)

572

Background: MET inhibition is emerging as a potent therapeutic strategy and MET gene amplification has shown predictive significance. MET amplification rate in mCRC, as previously reported in literature, varies from 9% in primary to 18% in metastases but intermixes increased copy number from chromosomal level aberrations with focal gene amplification. Validation of MET amplification rate in mCRC is needed. Methods: We performed analyses of MET amplification in mCRC patients (pts) (n = 636) across multiple cohorts. Cohort 1 (n = 103) included tissue microarray from liver metastases analysed using fluorescence in situ hybridization (FISH) [cMET and CEP7 probes, MET/CEP7 ratio > 2]. Cohort 2 (n = 205) included pts referred for phase I trials who had MET amplification testing using FISH. Cohort 3 (n = 279) included cases sequenced with HiSeq (Illumina) with full exome coverage for 202 genes including MET (average depth 800) with focal gene amplification (≥ 4 copies) identified by an in-house algorithm. Cohort 4 (n = 49) included pts refractory to EGFR monoclonal antibodies enrolled in the ATTACC (a prospective molecular screening) program for mCRC, in whom plasma circulating-free DNA (cfDNA) was analyzed by Guardant sequencing technology. Results: In tissue based analyses, focal MET amplification rate was 1.7% and was higher in primary tumors compared to metastases [3.1% (9/291) vs. 0.4% (1/288), p = 0.02] [Table]. In cohorts 2 & 3 MET amplification was found in 4 [MET/CEP7: 2.1 – 7.7; primary (4/130), metastases (0/75)] and 6 [copy number: 4.0 – 6.7; primary (5/161), metastases (1/110)] cases, respectively. MET amplification rate in pts who had progressed on anti-EGFR therapy was 14.3% (Table). Conclusions: Contrary to prior reports, in this large cohort, MET amplification was a rare event in mCRC pts and the rate was not higher in metastatic sites. However, MET amplification occurred in a sizable subset of pts refractory to anti-EGFR therapy as identified by cfDNA analysis. MET amplification appears to play a minor role in de novo colorectal carcinogenesis but may play an important role in acquired resistance to anti-EGFR therapy. [Table: see text]

Proteomic signatures of colorectal cancer to identify distinct and reproducible subgroups independent of oncogenic mutations

580

Background: While colorectal cancer (CRC) has classically been categorized on the basis of oncogenic mutations such as KRAS and BRAF, proteomic analyses directly elucidate the functional state of the cancer cell’s protein signaling, as recently described in a pan-cancer cohort and with mass-spectroscopy in a small CRC cohort. We performed an antibody-based proteomic analysis (reverse-phase protein array; RPPA) of a large cohort at MD Anderson (MDACC) and The Cancer Genome Atlas (TCGA) to determine patterns of protein expression in CRC. Methods: 725 archived CRC tumor samples (263 MDACC discovery set, and 462 TCGA validation set) underwent protein extraction and RPPA at MDACC to determine levels of 127 proteins. With unsupervised hierarchical clustering, samples dichotomized with distinct patterns of protein expression. The proteins with highest discriminatory utility were identified by LIMMA in the discovery set and confirmed in the validation set. Clinical variables and DNA sequencing results were available for correlation. Results: Among the top 30 discriminant proteins for the dichotomized groups in each dataset, 18 were common to both and tended to correlate with each other. One group was notable for high EMT (high fibronectin and collagen VI, low E-cadherin), while the other group was notable for high Akt/TSC/MTOR (high AKT, MTOR, Tuberin), and high RTK pathway components (high BRAF, HER2, HER3). This latter group also was notable for elevated beta-catenin and low CHK1, implicating differential activation of Wnt and cell cycle pathways, and intriguingly had elevated phospho-AMPK and phospho-NFkB. In the MDACC cohort, this latter group was more likely to have mucinous histology (p=0.009 by Fisher’s exact test) and lack lymphovascular invasion (p=0.026). When both TCGA and MDACC cohorts were examined, there was no significant difference in microsatellite instability, PIK3CA, KRAS, or BRAF mutations between the two proteomic groups. Conclusions: CRCs appear to be classifiable into distinct subsets by proteomic features. These findings reflect distinct differences in cellular signaling that are independent of common oncogenic driver mutations.

Proteomic features of colorectal cancer independently predict relapse-free survival

616

Background: Proteomic analysis continues to provide major insight into the pathophysiology of colorectal cancer (CRC). Recently, the Cancer Genome Atlas Project and others have utilized reverse-phase protein arrays (RPPAs) to identify critical protein markers and signaling pathways in a variety of tumor types. However, the prognostic relevance of many of these proteins remains unclear. We utilized RPPA to analyze stage 2 and 3 CRC to investigate the prognostic implications of the functional proteome. Methods: Protein extraction was performed on 232 snap frozen stage 2/3 samples from the MD Anderson Cancer Center with a median 5 year follow up. 163 validated proteins and phospho-proteins were analyzed by RPPA (with dichotomization by the median value), and used to identify protein predictors of tumor recurrence. Cox regression was used for univariate analysis with bootstrap validation, followed by inclusion of proteins with corrected p≤ 0.05 into multivariate model with clinical and pathologic variables, including stage, grade, and microsatellite status. Results: 12 proteins were found to be significant predictors of tumor recurrence on univariate analysis after bootstrap validation, which are notable for components of the energy balance/MTOR signaling pathway including AMPK, mTOR, PI3 Kinase p85, FoxO3a. On multivariate analysis, inclusive of known prognostic clinical and pathology variables, phospho-Bad (Ser112), FoxO3a, HER3, and phospho-S6 (Ser240-244) remained significant. Conclusions: Functional proteomic analysis has identified key proteomic features with prognostic importance independent of known clinical/pathological variables. Confirmation studies are ongoing to explore regulators of energy balance and apoptosis in colorectal cancer. [Table: see text]

Il-8 as an underutilized prognostic factor in metastatic colorectal cancer

409

Background: Cancer-associated inflammation has been identified as a key determinant of disease progression and survival in colorectal cancer. We investigated the association between circulating inflammatory cytokines and survival in metastatic colorectal cancer (mCRC) patients. Methods: Plasma levels of 47 cytokines were measured using multiplex-bead assays in a cohort of 168 previously untreated mCRC patients. Demographic, clinical-pathological features, body mass index, and mortality data were abstracted from patient medical records. Overall survival (OS) was evaluated by Kaplan-Meier analysis and Cox proportional hazards regression. Results: Using principal component analysis, we identified a subset of cytokines explaining the maximum variance in OS; and found interleukin (IL)-1b, IL-5, IL-8, IL-12 and VEGF to be significantly associated with OS. However, only IL-8 was significantly and independently associated with OS in multivariable-adjusted models. For each 100 pg/ml increase in the level of circulating IL-8, hazard rate for death increased by 1.6 (95% CI 1.24-1.97). IL-8 measurements ranged from <1 to 413 pg/ml with a median value of 22 pg/ml. Median uncensored survival was 26.5 and 15.5 months among patients with IL-8 levels below and above this value, respectively. ROC analysis of IL-8 demonstrated an AUC of 0.69 (95% CI 0.60-0.76), as compared to 0.52 for CEA (95% CI 0.46-0.59). Conclusions: We identified an association between IL-8 and OS in previously untreated mCRC patients, suggesting its potential role as a prognostic inflammatory biomarker. In this dataset, IL-8 outperformed CEA as a prognostic biomarker, a finding which requires validation in subsequent work. Appropriately identifying, monitoring and managing chronic inflammation and the host inflammatory response during colorectal cancer treatment may be important for improving long-term survival.

MET overexpression as a hallmark of the epithelial-mesenchymal transition (EMT) phenotype in colorectal cancer

3529

Background: Epithelial-mesenchymal transition (EMT) has been identified as a dominant molecular subtype of colorectal cancer (CRC). This EMT phenotype as recognized by complex gene signatures is prognostic and associated with chemoresistance, but a biomarker for EMT suitable for clinical utilization has not yet been validated. The purpose of this study was to compare MET protein expression with protein/gene expression of EMT markers and to evaluate its impact on overall survival (OS). Methods: We performed an exploratory analysis of 139 untreated primary CRC samples using data from The Cancer Genome Atlas. Protein and gene expressions were measured using reverse-phase protein array (RPPA) and RNA-sequencing, respectively. MET high/overexpressed group was defined by protein level in the highest quartile. Mann-Whitney U-test and Spearman rank correlation was used to determine association between MET protein expression and protein/gene expression of EMT markers and EMT gene signature scores. Regression tree method and Kaplan-Meier estimates were used to assess overall survival (OS). Results: The MET protein distribution is right skewed, demonstrating a unique population of MET high expressing tumors (P < 0.01). Colon tumors had higher MET protein levels compared to rectal tumors (P < 0.01). MET overexpression was associated with decreased OS (HR 2.92; 95% CI: 1.45 - 5.92). MET protein expression correlated strongly with protein expressions of SLUG (transcription factor for EMT) (r = 0.6) and ERCC1 (a marker for oxaliplatin chemo-resistance) (r = 0.6) (P < 0.01). Higher MET protein levels were associated with higher gene expression of 28 EMT markers including AXL, VIM, ZEB1, ZEB2, FGF1, TGFB1I1 and MMP11 (P < 0.05). Higher MET protein levels were also associated with higher gene scores derived from three published EMT gene signatures (P < 0.05). MET protein expression did not correlate with MET gene expression (r = 0.16). Conclusions: Increased MET protein expression strongly correlates with a molecular EMT phenotype and poor survival in patients with CRC. MET protein expression may be used as a surrogate biomarker to represent and select for this unique molecular subset of CRC driven by EMT biology.