Search

4079 Background: HER2 mutations are infrequent genomic events in biliary tract cancers (BTCs) and are associated with poor overall survival (OS) in patients with metastatic disease. HER2 overexpression is associated with an increased risk of disease recurrence in patients with resected BTC. There is limited data on targeting HER2 in BTC harboring activating somatic HER2 mutations. Neratinib, an irreversible, pan-HER, oral tyrosine kinase inhibitor, interferes with constitutive receptor kinase activation and has demonstrated activity in several HER2-mutant solid tumors. Methods: SUMMIT is an open-label, single-arm, multi-cohort, phase 2, ‘basket’ trial of neratinib in patients with solid tumors harboring oncogenic HER2 somatic mutations. The primary objective of the BTC cohort was to estimate objective response rate (ORR). Secondary objectives were clinical benefit rate (CBR), progression-free survival (PFS), OS, response duration, safety, and tolerability. Retrospective central confirmation of locally reported HER2 mutation (next-generation sequencing on archival or fresh tumor tissue using MSK-IMPACT or in cfDNA extracted from plasma by MSK-ACCESS) and association with outcome was an exploratory endpoint. This trial is registered with ClinicalTrials.gov (NCT01953926). Results: 25 treatment-refractory patients with metastatic BTC were enrolled (11 cholangiocarcinoma, 10 gallbladder, 4 ampullary cancers). ORR was 16% (95% CI 4.5–36.1%) and CBR was 28% (95% CI 12.1–49.4%). Median PFS and OS were 2.8 (95% CI 1.1–3.7) and 5.4 (95% CI 3.7–11.7) months, respectively. Median PFS for the gallbladder, cholangiocarcinoma and ampulla cohorts was 3.7 (95% CI 0.8–6.4), 1.4 (95% CI 0.5–9.1), and 1.1 (95% CI 1.1–3.8) months, respectively. Corresponding median OS values in these cohorts were 9.8 (95% CI 2.4–NE), 5.4 (95% CI 0.8–16.2), and 5.0 (95% CI 3.7–10.2) months, respectively. Central mutation confirmation was feasible for 23 of 25 patients; 22 were concordant with enrolment assays. The most common HER2 mutations were S310F (n = 11; 48%) and V777L (n = 4; 17%). Exploratory analyses suggested worse outcomes for HER2-mutant tumors with co-occurring oncogenic TP53 and CDKN2A alterations. Loss of amplified HER2 S310F and acquisition of multiple previously undetected oncogenic co-mutations were identified at progression in one of four responders. Diarrhea (56% any grade) was the most common toxicity. Conclusions: Neratinib is tolerable with modest antitumor activity in patients with BTC harboring HER2 mutations. Although the primary endpoint was met, future studies should evaluate rational combinations to augment and/or prolong responses. Clinical trial information: NCT01953926.

525 Background: Homologous recombination deficiency (HRD) status can be used to identify patients who are eligible for treatment with DNA damaging agents. Using a 3-biomarker Genomic Instability Score (GIS) threshold of ≥42, studies have previously examined the association between HRD status and outcomes in patients with triple negative breast cancer (TNBC). However, evidence suggests that a GIS threshold of ≥33 may be more appropriate. Here, we conducted an exploratory analysis evaluating the ability of ≥33 and ≥42 GIS thresholds to predict response to platinum-based treatment in patients with TNBC. Methods: Patients across 5 cohorts (TBCRC0301, TBCRC0082, NCT013725793, PrECOG 01054, combined cisplatin cohort4) were included in this analysis if they had a primary TNBC diagnosis, received neoadjuvant platinum-based treatment, had a valid GIS, and had known pathologic complete response (pCR) status. GIS was determined by a combination of loss of heterozygosity, telomeric-allelic imbalance, and large-scale state transitions.4,5 BRCA mutation status was defined by loss of function resulting from a pathogenic variant in BRCA1 or BRCA2. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), were calculated by comparing binary threshold status and binary pCR status. Results: A total of 204 tumors (158 BRCAwt; 33 BRCAm; 13 unknown) were included; pCR to platinum-based treatment occurred in 55 cases (39 BRCAwt; 14 BRCAm; 2 unknown). Sensitivity, specificity, PPV, and NPV were comparable between the ≥33 and ≥42 GIS thresholds, with the ≥33 threshold producing higher sensitivity values. This was true when thresholds were applied to all samples and to BRCAwt samples only (Table). Among patients who achieved pCR in response to platinum-based treatment, 5.5% of patients in the full cohort and 7.7% of those in the BRCAwt cohort had a GIS between 33-41. Conclusions: To ensure that the majority of patients likely to benefit from treatment are identified, a GIS of ≥33 may be the most appropriate threshold to predict response to platinum-based treatment in patients with TNBC; however, a prospective trial will be needed to confirm these findings. Additional studies will be important to determine whether this threshold may be appropriate to determine eligibility for other DNA-damaging agents such as PARP inhibitors. 1. Ann Oncol. 2020;31(11):1518-25 2. J Nucl Med. 2015;56(1):31-7. 3. Breast Cancer Res Treat. 2015;151(3):629-38. 4. Clin Cancer Res. 2016;22(15):3764-73. 5. Breast Cancer Res Treat. 2014;16(6):1-9. [Table: see text]

3003 Background: NRG1 fusion proteins are oncogenic drivers in pancreas cancer and other solid tumors. They bind HER3, leading to HER2/HER3 heterodimerization and oncogenic transformation. The activity of zenocutuzumab (MCLA-128; zeno), a bispecific antibody targeting NRG1 fusion signaling in NRG1 fusion positive ( NRG1+) cancers, is being evaluated in the ongoing global multicenter phase 2 part of the eNRGy study and a global early access program (EAP). Methods: Enrolled patients (pts) have advanced NRG1+ pancreas cancer, non-small cell lung cancer (NSCLC), and other solid tumors previously treated with standard therapy, are ≥ 18 years-old, have ECOG ≤1, adequate organ function, and measurable disease (RECIST v1.1). Zeno dosing: 750 mg IV every 2 weeks until progression or unacceptable toxicity. Primary endpoint: investigator (INV)-assessed objective response rate (ORR). Secondary endpoints: ORR per central independent radiologist review, duration of response (DOR), and safety. Tumor imaging is conducted every 8 weeks. Results: 51 pts with NRG1+ cancer have received zeno, 37 in the eNRGy study and 14 pts in the EAP. As of 12 Jan 2021, treatment is ongoing in 27/51 pts (8/13 pancreas, 10/25 NSCLC, 9/13 other solid tumors). Among the 51 pts, 10 pts with pancreas cancer, 18 pts with NSCLC, and 5 pts with other solid tumors had measurable disease and had the opportunity for ≥1 tumor assessment (TA) and are included in this analysis. Among the 10 pts with pancreas cancer, median age was 49 y (range 34-72), 50% were male, 6/4 pts had ECOG 0/1, and all had metastatic disease and were KRAS wild-type. The median number of prior therapies was 3 (range 1-6). The INV-assessed confirmed ORR was 40% (4/10; 90% CI, 15;70), and for this cohort of pts, responses occurred at the first TA. Tumor regression was seen in 7/10 pts, and the disease control rate was 90% (90% CI, 61-100). A CA 19-9 decline of ≥ 50% was observed in 9/9 (100%) pts. DOR is pending. In the overall NRG1+ population, tumor regression was observed in 25 of 33 pts and confirmed INV-assessed responses were seen in 9 of 33 pts (ORR 27%; 90% CI, 15;43), including in pts who previously received afatinib. Zeno was well tolerated with no pts requiring dose reduction for toxicity. Across all cohorts, for individual AEs, grade 3 events were reported in ≤5% of pts, and there was a notable lack of cardiotoxicity and severe gastrointestinal or skin toxicity. Updated data from all cohorts (pancreas, NSCLC, other solid tumors) will be presented. Conclusions: Zeno induces rapid and major radiologic tumor regression and biomarker responses in heavily-pretreated metastatic KRAS wild-type NRG1+ pancreas cancer, with minimal toxicity. Zeno is a promising novel targeted therapeutic option for pts with NRG1+ cancers. Clinical trial information: NCT02912949.

TPS1109 Background: PALB2 (Partner and Localizer of BRCA2) plays a critical role in the maintenance of genomic integrity through its role in the Fanconi anemia and homologous recombination DNA repair pathways. Our recently reported phase II clinical trial (Gruber, et al. JCO 2019) showed that talazoparib monotherapy was associated with single agent activity in germline PALB2 (gPALB2) mutation-associated advanced breast cancers. Of 5 patients with a germline PALB2 mutation, all 5 had reduction in target lesions > 20% with 3 of 5 achieving a RECIST 1.1 response. All patients had visceral metastases and were heavily pretreated; one response occurred in a patient with 8 prior lines of therapy for metastatic breast cancer. Clinical activity of the PARP inhibitor olaparib was also demonstrated in breast cancer patients with gPALB2 mutations in the Olaparib Expanded trial (TBCRC 048). Thus, we believe there is significant value in generating additional PARP inhibitor monotherapy data in subjects with advanced PALB2 mutation-associated breast cancer. We propose this phase II trial to better estimate the rate of RECIST 1.1 objective response in this patient population. Methods: This is a single-arm, two-stage, non-randomized study to assess the objective response rate (ORR) of talazoparib monotherapy in patients with PALB2 mutation-associated advanced breast cancer. Eligible subjects will be adults with inoperable locally advanced or metastatic breast cancer with a germline or somatic PALB2 mutation with 0-3 prior therapies for advanced disease. Eligible patients must have a deleterious or suspected deleterious mutation in PALB2 on a CLIA approved commercial germline or next generation sequencing tumor assay. Study treatment is talazoparib 1 mg PO daily. In stage 1, ORR will be assessed after 15 patients; if at least 2 responses are observed then an additional 15 patients will be enrolled; if less than 2 responses are observed the study will be terminated. The primary objective is to evaluate whether talazoparib monotherapy can induce a 30% ORR in subjects with advanced breast cancer associated with a PALB2 mutation. Response will be assessed by RECIST 1.1. Secondary objectives include safety, PFS, clinical benefit rate, and patient-reported quality-of-life. Exploratory endpoints will assess for the presence of PALB2 loss-of-heterozygosity, biallelic mutations, correlation of ctDNA profile with treatment response and the correlation of germline versus somatic mutations with response rate. This trial is currently activated and enrolling at Stanford Cancer Center and it is expected that 4 additional sites will be added. Conclusion: This trial will evaluate gPALB2 as a biomarker for PARP inhibitor monotherapy in advanced or metastatic HER2- breast cancer. Clinical trial information: pending .

320 Background: Genomic profiling studies have reported somatic HER2 mutations in ~2–5% of biliary tract cancers (BTC). Clinical data from the SUMMIT study demonstrate that neratinib, a pan-HER irreversible tyrosine kinase inhibitor, has encouraging clinical activity in multiple types of HER2-mutant solid tumor malignancies. Methods: SUMMIT is a multi-histology, open-label, phase II ‘basket’ study of neratinib in patients with somatic HER2 mutations (ClinicalTrials.gov NCT01953926). Patients with activating somatic HER2 mutations with different histologies, including BTC, received neratinib monotherapy (240 mg oral daily). Loperamide prophylaxis was mandatory during cycle 1. Efficacy endpoints: objective response rate (ORR, RECIST v1.1); clinical benefit rate (CBR); duration of response; progression-free survival (PFS). Adverse events (AEs) were assessed by CTCAE v4.0. Genomic profiling from fresh/archival tumor tissues and/or plasma cfDNA was performed retrospectively by next-generation sequencing (MSK-IMPACT). Results: As of 3-Sep-2020, 25 patients with HER2-mutant BTC were enrolled: gallbladder (40%); intrahepatic (24%); extrahepatic (20%); ampulla of Vater (16%). 68% of patients received ≥2 systemic regimens (96% received prior gemcitabine-based regimens). The S310F/Y variant accounted for nearly half of HER2 mutations (n=11). Other HER2 mutations: V777L (n=5); L755S (n=2); V842I (n=2); R678Q (n=2). Confirmed ORR in 25 evaluable patients was 12% (95% CI 3–31%) and CBR was 20% (95% CI 7–41%), including 3 confirmed PRs and 2 patients with SD for ≥16 weeks. Tumor shrinkage was observed in multiple HER2-activating mutations and enriched in gallbladder and extrahepatic subtypes of BTC. Median PFS was 2.8 (95% CI 1.1–3.7) months; median overall survival (OS) was 5.4 (95% CI 3.7–11.7) months. Nine (36%) patients (3 of whom with ECOG PS 2) came off study within 28 (range 6–47) days of treatment due to clinical deterioration (unrelated to study drug) followed by death. The most common treatment-related AEs (any grade) were diarrhea (56%) and vomiting (48%). Diarrhea was the most common Grade 3 event (24%); 4 patients (16%) required a neratinib dose reduction; no patients discontinued treatment due to diarrhea. Conclusions: Neratinib is safe and tolerable in patients with advanced BTC patients and somatic HER2 mutations. The antitumor activity of neratinib appears comparable to current standards of care, with similar PFS and OS in heavily pretreated patients. Analysis of co-occurring oncogenic mutations and response is ongoing, and consideration is being given to neratinib-based combination regimens to further improve outcomes in this setting. Clinical trial information: NCT01953926.

1037 Background: MCLA-128 (zenocutuzumab) is an ADCC-enhanced humanized bispecific antibody targeting HER2 and HER3 and potently blocking HER3-ligand induced receptor dimerization. Upregulation of Her2:Her 3 pathway is a means of resistance to ET in HR+ breast cancer, indicating a potential role for MCLA-128. In preclinical studies, the combination of MCLA-128 with ET in breast cancer xenografts outperformed single drug treatments. The current study explores the use of MCLA-128 to rescue pts with ET-resistant MBC who have progressed on a CDK4/6i. Methods: This phase II, open-label trial planned for up to 40 evaluable women with HR+, HER2 low (IHC 1+/IHC 2+ with negative FISH) MBC, who had progressed on a CDK4/6i and up to 3 lines of ET, who had received ≤ 2 chemotherapy regimens in the metastatic setting. Pts received MCLA-128 (750 mg, 2h IV, flat dose) q3w combined with last ET on which the pt had previously progressed immediately prior to study entry. Disease control rate (DCR; RECIST 1.1, per investigator), best overall response (BOR), overall response rate (ORR), safety, and PK, are evaluated. Data cut off was 14Nov2019. Results: 48 pts were treated, all of whom had progressed on a CDK4/6i. Pts had received a median 2 prior ET lines (range 1-5) and 1 line (range 1-3) of chemotherapy. Pts had a median number of 3 metastatic sites (range 1-6) and 42 (88%) had visceral involvement. Among 42 pts evaluable for efficacy, DCR was 45% (90% CI 32-59) with 2 pts having unconfirmed PR and 19 pts SD as BOR. Common related AEs (all grades; G3-4) were asthenia/fatigue (27%; 2%), diarrhea (25%; 0), nausea (21%; 0). No clinically significant LVEF decline was seen. At the end of cycle 1, mean trough level of MCLA-128 was 15.5 µg/mL, and mean terminal half-live was 102 h (n = 19-21). Data on the primary endpoint, clinical benefit rate at 24 weeks, and biomarkers will be provided. Conclusions: The addition of MCLA-128 to the last line of ET showed clinical activity after ET+CDK4/6i failure and a favorable safety profile. Clinical trial information: NCT03321981 .

1549 Background: While germline pathogenic variants (PVs) in BRCA1/2 account for a large proportion of hereditary breast cancer (BC), PVs in PALB2, CHEK2 and ATM are increasingly detected. However, the phenotype and clinical features of invasive BC with these PVs have not been fully described. Methods: We identified patients with a PV or likely PV in PALB2, CHEK2 or ATM tested clinically at Stanford between 2014 - 2019 who provided informed consent to be included in a prospective cancer genetics registry. Data on baseline demographics, genetic testing history, and clinicopathologic features of diagnosed BC were collected. For patients with a subsequent diagnosis of metastatic BC, we calculated disease-free interval (DFI). Results: 130 patients met inclusion criteria for analysis: ATM (N=39), CHEK2 (N=58), PALB2 (N=33). Nearly all (98.5%) were women, with 2 male BC in ATM carriers. Non-Hispanic White ethnicity was most common in ATM (64.1%, 95% CI 48.4%-77.3%) and CHEK2 carriers (69.0%, 95% CI 56.1%-79.4%), but comprised only 39.4% (95% CI 24.7%-56.4%) in PALB2 carriers. Asian/Pacific Islander (24.2%, 95% CI 12.6%-41.3%) and Hispanic (30.3%, 95% CI 17.3%-47.5%) ethnicities were enriched among PALB2 mutation carriers. In total, 97.7% learned of their PV status only after a preceding diagnosis of BC and 43.1% were diagnosed with BC at age ≤ 45. Data regarding invasive BC subtypes, incidence of subsequent primary BC, and metastatic recurrence are listed below in the table. Additional data on stage, grade and sites of metastatic spread will be presented. Conclusions: We observed clinically important differences in the spectrum of BC subtypes among carriers of ATM, CHEK2 and PALB2 PVs, in addition to racial/ethnic differences with Asian/Pacific Islander and Hispanic ethnicity enriched among carriers of PALB2 PVs. [Table: see text]

432 Background: Gastric cancer (GC) is the fifth leading cancer diagnosis and third frequent cause of cancer death globally. GC results from a cascade of molecular and genetic changes owing to environmental and genetic factors. While there are known risks to GC, more is to be learned about its development so to establish effective screens for early stages of cancer. Methods: The Gastric Cancer Registry (GCR) was built to investigate GC and discover informative genomic biomarkers. The GCR comprises medical, family, and social history and genomics data from patients with GC, a family history of GC, and/or a known mutation in the gene CDH1. Samples of saliva and gastric tumors are collected when available. The GCR allows us to leverage several genome sequencing datasets to construct a complete molecular landscape of GC. Early analysis of GC tissue includes whole exome sequencing (WES) to identify mutations, whole genome sequencing (WGS) for copy number variation, and RNA sequencing (RNAseq) for expression profiling. Results are used to inform of clonal neoantigens, microbiome, and immune cell populations. Saliva will be analyzed with linked reads sequencing to unearth germline mutations not picked up in standard clinical gene panels. Results: Datasets from 455 patients and samples from 159 patients have been collected. In a pilot study, we pinpointed specific mutations using WES and revealed extensive changes in genome copy number involving clinically actionable genes through WGS. Most tumors had a high degree of genomic instability and exhibited candidate markers for treatment decisions and response. Additionally, we found that RNAseq revealed tumor subgroups through gene expression signatures. Conclusions: The GCR is an informative resource enabling the identification of biomarkers for GC. With the GCR we are integrating expertise in translational cancer genomics with molecular biology, statistics, and bioinformatics to build a platform for discovery and the development of tools that will ultimately improve the detection, treatment, and prevention of GC.

The American Society of Clinical Oncology (ASCO) has long affirmed that the recognition and management of individuals with an inherited susceptibility to cancer are core elements of oncology care. ASCO released its first statement on genetic testing in 1996 and updated that statement in 2003 and 2010 in response to developments in the field. In 2014, the Cancer Prevention and Ethics Committees of ASCO commissioned another update to reflect the impact of advances in this area on oncology practice. In particular, there was an interest in addressing the opportunities and challenges arising from the application of massively parallel sequencing—also known as next-generation sequencing—to cancer susceptibility testing. This technology introduces a new level of complexity into the practice of cancer risk assessment and management, requiring renewed effort on the part of ASCO to ensure that those providing care to patients with cancer receive the necessary education to use this new technology in the most effective, beneficial manner. The purpose of this statement is to explore the challenges of new and emerging technologies in cancer genetics and provide recommendations to ensure their optimal deployment in oncology practice. Specifically, the statement makes recommendations in the following areas: germline implications of somatic mutation profiling, multigene panel testing for cancer susceptibility, quality assurance in genetic testing, education of oncology professionals, and access to cancer genetic services.

Purpose This study was designed to assess efficacy, safety, and predictors of response to iniparib in combination with gemcitabine and carboplatin in early-stage triple-negative and BRCA1/2 mutation–associated breast cancer. Patients and Methods This single-arm phase II study enrolled patients with stage I to IIIA (T ≥ 1 cm) estrogen receptor–negative (≤ 5%), progesterone receptor–negative (≤ 5%), and human epidermal growth factor receptor 2–negative or BRCA1/2 mutation–associated breast cancer. Neoadjuvant gemcitabine (1,000 mg/m2 intravenously [IV] on days 1 and 8), carboplatin (area under curve of 2 IV on days 1 and 8), and iniparib (5.6 mg/kg IV on days 1, 4, 8, and 11) were administered every 21 days for four cycles, until the protocol was amended to six cycles. The primary end point was pathologic complete response (no invasive carcinoma in breast or axilla). All patients underwent comprehensive BRCA1/2 genotyping, and homologous recombination deficiency was assessed by loss of heterozygosity (HRD-LOH) in pretreatment core breast biopsies. Results Among 80 patients, median age was 48 years; 19 patients (24%) had germline BRCA1 or BRCA2 mutations; clinical stage was I (13%), IIA (36%), IIB (36%), and IIIA (15%). Overall pathologic complete response rate in the intent-to-treat population (n = 80) was 36% (90% CI, 27 to 46). Mean HRD-LOH scores were higher in responders compared with nonresponders (P = .02) and remained significant when BRCA1/2 germline mutations carriers were excluded (P = .021). Conclusion Preoperative combination of gemcitabine, carboplatin, and iniparib is active in the treatment of early-stage triple-negative and BRCA1/2 mutation–associated breast cancer. The HRD-LOH assay was able to identify patients with sporadic triple-negative breast cancer lacking a BRCA1/2 mutation, but with an elevated HRD-LOH score, who achieved a favorable pathologic response. Confirmatory controlled trials are warranted.

3006 Background: Talazoparib, a PARP inhibitor, is active in germline BRCA1/2 mutant advanced HER2-negative breast cancer, but its activity beyond BRCA1/2 is unknown. We conducted a single institution phase II trial to evaluate talazoparib in patients (pts) with advanced HER2-negative breast cancer or other solid tumors with a germline (g) or somatic (s) alteration in HR pathway genes not including BRCA1/2. Methods: Eligible pts had measurable disease, lacked a germline or somatic mutation in BRCA1/2, received at least one prior therapy for advanced HER2-negative breast cancer or other solid tumor and had a HR pathway gene mutation: PALB2, CHEK2, ATM, NBN, BARD1, BRIP1, RAD50, RAD51C, RAD51D, MRE11, ATR, PTEN, FANCA, FANCC, FANCD2, FANCE, FANCF, FANCG, FANCL. Pts with no progression on or within 8 weeks of their last platinum dose were eligible. Pts were treated with talazoparib 1 mg po daily until disease progression. Response was assessed every 8 +/- 1 weeks. If 2 or more responses were observed in 10 pts in stage I, the study would proceed to stage II and enroll 10 additional pts. The null hypothesis of a ≤ 5% objective response rate would be rejected if at least 3 of 20 respond. Results: Twenty pts were enrolled; 13 breast cancer (12 HR+/HER2-, 1 TNBC) and 7 non-breast cancer (pancreas, colon, uterine, testicular, parotid salivary). Median age was 54 years. Of 12 response evaluable pts with breast cancer, 3 had a RECIST response (ORR = 25%, 2 g PALB2, 1 g CHEK2/g FANCA/sPTEN) and 3 additional pts (g PALB2, s ATR, s PTEN) had SD ≥ 6 months (CBR = 50%). No responses were seen in non-breast tumors; 2 (g CHEK2 testicular, g ATM colon) had SD ≥ 6 months. Talazoparib was well tolerated; 5 patients required dose reduction for hematologic toxicity. Results of tumor HR deficiency status assessment from metastatic biopsies and serial ctDNA profiling will be presented. Conclusions: In this proof-of-concept phase II study, single agent talazoparib demonstrated activity in HER2-negative advanced breast cancer pts with a HR pathway mutation beyond BRCA1/2. Further evaluation of talazoparib in this population is warranted. Clinical trial information: NCT02401347.

1525 Background: Guidelines recommend consideration of prophylactic surgery for patients with a germline pathogenic variant in some cancer predisposition genes. We assessed surgery utilization in a prospective, multi-institutional cohort study of MGPT. Methods: 2000 patients had MGPT and completed questionnaires at 3, 6, and 12 months. Patients reported surgical utilization and indication (treatment or prevention). Surgery utilization was assessed according to cancer history and MGPT test results: Positive, pathogenic variant; VUS, variant of uncertain significance; Negative, benign variants. Results: Overall, 12.9% (198/1537) of patients reported surgery after MGPT (median follow-up 13 months). Only 31.3% (62/198) of patients specified that their surgery was preventive. Preventive surgery utilization was significantly higher among patients who tested positive (n=30, 14.9%) compared to those testing negative (n=20, 2.3%, p

215 Background: In light of recent FDA approval of anti-PD-1 therapy for microsatellite unstable (MSI-H) or mismatch repair deficient (dMMR) solid tumors, identifying patients with dMMR tumors has become increasingly important. While screening for dMMR and MSI-H in colorectal cancer (CRC) is recommended, this screening is less commonly done for extracolonic gastrointestinal (GI) tumors. At Stanford Comprehensive Cancer Institute (SCCI), all GI cancer patients have been universally screened for dMMR via immunohistochemistry since January 2016. Methods: In this study, we undertook a retrospective review of all GI cancer patients screened for dMMR between January 2016 to December 2017. Data on patient characteristics, germline and tumor sequencing, and tumor characteristics were collected and reported. Results: A total of 1543 GI malignancies were screened for dMMR at SCCI during the study period. Colorectal (n = 711), pancreatic (n = 264), gastric (n = 159), and gastro-esophageal (n = 137) cancer were amongst the most frequently screened tumors. dMMR was detected in 7.1% of all GI malignancies. We detected the highest prevalence of dMMR in colorectal (69/711, 9.7%), followed by gastric (15/159, 9.4%), pancreatic (18/264, 6.8%), and gastro-esophageal cancer (6/137, 4.4%). Lynch syndrome was the most common etiology for dMMR in CRC patients (39.4%), double somatic (confirmed or possibly) mutations were most common in pancreatic cancer (44.4%), and somatic MLH1 hypermethylation was the most common etiology in gastric (73.3%) and gastro-esophageal cancer (100%). Conclusions: Given the relatively high incidence of dMMR in GI malignancies, we strongly recommend screening all GI malignancies. Of note, we found higher than previously reported rates of dMMR within pancreatic cancer. Our results also suggest that while rare, double somatic mismatch repair mutations may be a significant pathway causing dMMR in pancreatic cancer.

Purpose In the United States, gastric cancer is rapidly fatal with a 25% 5-year survival. Of the few patients who survive, little is known about their demographic, clinical, and tumor characteristics. Patients and Methods Data regarding all cases of gastric and gastroesophageal junction (GEJ) adenocarcinoma diagnosed in California between 1988 and 2005 were obtained from the California Cancer Registry, a member of the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program. A Cox proportional hazards model was constructed to understand the independent relationships of patient demographic, disease, and treatment factors with survival. Results We identified 47,647 patients diagnosed with gastric or GEJ cancer. Of those, only 9,325 (20%) survived at least 3 years. Variables associated with longer survival were localized stage (hazard ratio [HR], 0.20), surgery with diagnosis in 2002 or later (HR, 0.34), surgery with diagnosis in 2001 or before (0.37), regional stage (HR, 0.53), chemotherapy (HR, 0.56), intestinal histology (HR, 0.74), well- or moderately differentiated tumors (HR, 0.76), radiation (HR, 0.80), Asian/Pacific Islander race (HR, 0.81), treatment at an academic hospital (HR, 0.85), fundus/body/antrum location (HR, 0.90), highest socioeconomic status quintile (HR, 0.91), female sex (HR, 0.92), Hispanic race (HR, 0.92), and hospital size more than 150 beds (HR, 0.94). Kaplan-Meier curves showed longer median disease-specific survival (DSS) in patients with tumors originating in the fundus/body/antrum compared with esophagus/cardia (13.4 v 10.8 months). Intestinal histology had significantly longer median DSS (28.9 months) compared with other (11.0 months) or diffuse (10.1 months) histology. Conclusion Patients who survive gastric and GEJ cancer more than 3 years after diagnosis have demographic and pathologic characteristics distinct from those who do not survive.

1582Background: Cancer screening guidelines recommend that germline carriers with a pathogenic variant (PV) in a CRC susceptibility gene should undergo more frequent colonoscopy screening. We asses...

1581Background: Cancer screening guidelines recommend that germline carriers with a pathogenic variant (PV) in a breast cancer susceptibility gene undergo more intensive breast screening including ...

Purpose Lynch syndrome is associated with inherited germline mutations in mismatch repair (MMR) genes. Genetic testing in high-risk individuals may yield indeterminate results if no mutation is found or if a mutation of unclear pathogenic significance is observed. There are limited data regarding how well patients with Lynch syndrome understand the clinical implications of genetic test results. This study examines colorectal cancer (CRC) risk perception in individuals tested for MMR mutations and identifies the factors associated with an appropriate interpretation of their cancer risk. Patients and Methods A total of 159 individuals who met the Revised Bethesda Guidelines and had previously undergone genetic testing completed a questionnaire eliciting demographic data, cancer history, genetic test results, and an estimate of their CRC risk. Associations between clinical factors, genetic test results, and CRC risk perception were explored using multivariable analyses. Results Of the 100 individuals with a pathogenic mutation (true positive), 90 (90%) correctly estimated their CRC risk as “high” or “very high” compared with other individuals their age. However, only 23 (62%) of 37 individuals with an indeterminate genetic test result correctly estimated their risk. Individuals with a history of Lynch syndrome–associated cancer (odds ratio [OR], 0.1; 95% CI, 0.1 to 0.6) or indeterminate genetic test results (OR, 0.2; 95% CI, 0.1 to 0.6) were significantly less likely to estimate their CRC risk as increased. Conclusion Patients at risk for Lynch syndrome with an indeterminate genetic test result may be falsely reassured. It is important that health care providers continue to discuss the implications of uninformative results on lifetime cancer risk.

Purpose To investigate the gefitinib, fluorouracil (FU), leucovorin, and oxaliplatin regimen (IFOX) in previously treated patients with metastatic colorectal cancer. Patients and Methods Eligible patients had stage IV colorectal adenocarcinoma and had demonstrated progression or intolerance to a prior chemotherapy regimen not including oxaliplatin. Each cycle consisted of 14 days. Cycle 1 consisted of oxaliplatin 85 mg/m2 intravenously (IV) during 2 hours on day 1, hours 0 to 2; leucovorin 200 mg/m2 IV on days 1 and 2, hours 0 to 2; FU 400 mg/m2 IV push on days 1 and 2; and FU 600 mg/m2 IV on days 1 and 2, hours 2 to 24 (FOLFOX-4). All subsequent cycles consisted of FOLFOX-4 with gefitinib at 500 mg/d administered orally throughout the 14-day cycle. Results Twenty-seven patients were enrolled onto the study. The median number of prior chemotherapy regimens was two, and 74% of all patients received prior irinotecan. Nine of the 27 patients (33%) and six of the 20 patients (30%) who had prior FU and irinotecan had a partial response by Response Evaluation Criteria in Solid Tumors Group criteria. Median overall survival was 12.0 months. Median event-free survival was 5.4 months. Grade 3 to 4 toxicities included neutropenia (48%), diarrhea (48%), nausea (22%), and vomiting (15%). Conclusion IFOX is an active regimen in patients with previously treated metastatic colorectal adenocarcinoma, demonstrating higher response rates than those reported with FOLFOX-4 alone in a similar patient population.

1525 Background: Genetic testing is a powerful tool for stratifying cancer risk. Multiplex gene panel (MGP) testing allows simultaneous analysis of multiple high- and moderate- penetrance genes. However, the diagnostic yield and clinical utility of panels remain to be further delineated. Methods: A report of a fully accrued trial (N = 2000) of patients undergoing cancer-risk assessment. Patients were enrolled in a multicenter prospective cohort study where diagnostic yield and off-target mutation detection was evaluated of a 25 gene MGP comprised of APC, ATM, BARD1, BMPR1A, BRCA1, BRCA2, BRIP1, CDH1, CDK4, CDKN2A, CHEK2, EPCAM, MLH1, MSH2, MSH6, MUTYH, NBN, PALB2, PMS2, PTEN, RAD51C, RAD51D, SMAD4, STK11, TP53. Patients were enrolled if they met standard testing guidelines or were predicted to have a ≥2.5% mutation probability by validated models. Differential diagnoses (DDx) were generated after expert clinical genetics assessment, formulating up to 8 inherited cancer syndromes ranked by estimated likelihood. Results: 1998/2000 patients had reported MGP test results. Women constituted 81% of the sample, and 40% were Hispanic; 241 tested positive for at least 1 pathogenic mutation (12.1%) and 689 (34.5%) patients carried at least 1 variant of uncertain significance. The most frequently identified mutations were in BRCA1 (17%, n = 41), BRCA2 (15%, n = 36), APC (8%, n = 19), CHEK2 (7%, n = 17), ATM (7%, n = 16). 39 patients (16%) had at least 1 pathogenic mutation in a mismatch repair (MMR) gene ( MLH1, n = 10; MSH2, n = 10; MSH6, n = 8; PMS2, n = 11). 43 individuals (18%) had MUTYH mutations – 41 were monoallelic. Among 19 patients who had mutations in APC – 16 were APC I1307K. Only 65% (n = 159) of PV results were included in the DDx, with 35% (n = 86) of mutations not clinically suspected. Conclusions: In a diverse cohort, multiplex panel use increased genetic testing yield substantially: 35% carried pathogenic mutations in unsuspected genes, suggesting a significant contribution of expanded multiplex testing to clinical cancer risk assessment. The identification of off-target mutations broadens our understanding of cancer risk and genotype-phenotype correlations. Follow-up is ongoing to assess the clinical utility of multiplex gene panel testing. Clinical trial information: NCT02324062.

1523 Background: Mutation carrier prediction models are clinically useful tools for identifying candidates for genetic counseling and testing. Consensus guidelines recommend germline genetic testing for those with a carrier probability (CP) of approximately 5% or higher. However, prediction models may perform less well among racial/ethnic minorities. Our hypothesis is that pathogenic mutations (PM) are identifiable in a clinically meaningful fraction of racially/ethnically diverse patients with a CP of < 5%. Methods: We conducted a multicenter prospective clinical trial of patients undergoing cancer-risk assessment using a 25 gene panel, which include APC, ATM, BARD1, BMPR1A, BRCA1, BRCA2, BRIP1, CDH1, CDK4, CDKN2A, CHEK2, EPCAM, MLH1, MSH2, MSH6, MUTYH, NBN, PALB2, PMS2, PTEN, RAD51C, RAD51D, SMAD4, STK11, TP53. Patients were recruited from August 2014 to November 2016 at three centers. Patients were enrolled if they met standard clinical criteria for genetic testing or were predicted to have a ≥2.5% probability of inherited cancer susceptibility using validated prediction models. We evaluated the CP of patients with a PM in BRCA1, BRCA2, and/or a mismatch repair (MMR) gene using the following models: (1) BRCApro, (2) MMRpro and (3) PREMM1,2,6. Results: Of 2000 patients enrolled in this cohort, 80.6% are female (n = 1612). Regarding race/ethnicity, the cohort is 40.1% Non-Hispanic White (n = 802), 37.4% Hispanic (n = 748), 11.5% Asian (n = 230), 3.9% Black (n = 78), and 7.1% Other (n = 142). Among 241 (12.1%) patients who tested positive for a pathogenic mutation, 76 (31.5%) patients had a BRCA1 or BRCA2 mutation. Of those, 52 (68.4%) patients had a BRCApro CP of < 5%. Thirty-eight (15.8%) patients had a pathogenic mutation in an MMR gene: 19 (50.0%) had an MMRpro CP of < 5%, while 13 (34.2%) had a PREMM1,2,6 CP of < 5%. The racial/ethnic distribution of BRCA1/2 or MMR mutation carriers is similar to that of the whole cohort. Conclusions: In a diverse cohort of patients undergoing 25-gene multiple-gene panel testing, half or more carriers of BRCA1/2 or MMR mutations had a CP of < 5%, the consensus guideline-recommended cutoff for genetic testing. These results support a lower threshold for genetic testing guidelines. Clinical trial information: NCT02324062.

1576 Background: Sequencing more genes increases the chance of finding a pathogenic mutation and/or a variant of uncertain significance (VUS). Little is known about potential harms of multiplex testing for cancer risk, such as unwarranted surgery or adverse psychological effects. Methods: We conducted a prospective trial of sequencing 25 genes : APC, ATM, BARD1, BMPR1A, BRCA1, BRCA2, BRIP1, CDH1, CDK4, CDKN2A, CHEK2, EPCAM, MLH1, MSH2, MSH6, MUTYH, NBN, PALB2, PMS2, PTEN, RAD51C, RAD51D, SMAD4, STK11, TP53. Patients were eligible if they met standard testing guidelines or predictive models estimated ≥2.5% mutation probability. Participants were surveyed 3 months post-test: the Multidimensional Impact of Cancer Risk Assessment (MICRA) scale measured distress, uncertainty and positive experiences. We report on the fully accrued trial (N = 2000). Results: 1998/2000 (99.9%) participants currently have reported results: 12.1% tested positive for a pathogenic mutation (Pos), 34.5% had VUS only and 53.5% tested negative (Neg). Median age was 51, 81% were female, 40% Hispanic, and 72% had a cancer history. Self-reported preventive surgery rates were low (mastectomy 9.3%, hysterectomy 1.5%, oophorectomy 1.6%), with no difference between VUS and Neg patients (p = 0.346). Most patients never or rarely had thoughts of cancer affecting daily activities (Pos 59.5%, VUS 66.9%, Neg 71.0%), never regretted testing (Pos 84.1%, VUS 90.0%, Neg 93.6%), and wanted to know all results, even those that doctors do not fully understand (Pos 81.7%, VUS 78.8%, Neg 77.1%). Pos patients had higher MICRA distress and uncertainty scores than VUS and Neg patients, whose distress and uncertainty scores did not differ significantly (p = 0.165, p = 0.129). Relatives of Pos patients completed genetic testing (30.4%) more often than VUS (5.8%) or Neg patients (5.1%, p < 0.001). Conclusions: After multiplex testing of 2000 diverse patients, few reported preventive surgery at 3 months; VUS patients had no more distress, regret or uncertainty than Neg patients. Pos patients most often advised relatives to test, suggesting that participants understood the implications of test results. Longer-term follow-up of test-related outcomes is underway. Clinical trial information: NCT02324062.

1528 Background: Next-generation sequencing technology enables more comprehensive germline genetic testing, including genes whose cancer risks are less well-characterized (particularly among patients with less striking family histories). Little is known about patient outcomes, particularly adherence to risk-reducing recommendations and family testing. Methods: We attempted a phone interview ≥3 times with each adult patient who had a germline pathogenic or likely pathogenic mutation found in an emerging cancer gene (defined as any gene other than BRCA1/2 or the Lynch Syndrome genes MLH1, MSH2/6, PMS2) at a single academic cancer genetics clinic from January 2013-July 2016. Results: Of 143 eligible patients, 53 (37%) were successfully contacted and all consented to participate. Median follow-up was 677 days (range 247–1401) and age was 52 years (21-82). Two-thirds (68%) had personal cancer history and 93% had a first-degree relative with cancer. Mutations in genes associated with named syndromes ( APC=5 , CDH1=3 , TP53=3, PTEN=2) were found in 23% (many of whom lacked family history typical of these syndromes) whereas 77% had mutations in less well-characterized genes ( MYH=10, CHEK2=10, ATM=6, PALB2=6 , NBN=3, RAD51C=2, RAD51D=2, SDHB=1, CDKN2A=1, MRE11A=1, RAD50=1, FLCN=1). Conclusions: Two years after panel testing, patients with germline mutations in less well-characterized genes reported high rates of adherence to recommendations, family communication and testing. Limitations include a relatively low response rate and a single academic center; this may bias toward a “best-case” scenario. Larger, population-based studies will be crucial to understand the real-world outcomes of germline multiple-gene panel testing and its contribution to precision oncology. [Table: see text]

PURPOSE: Motexafin gadolinium is a magnetic resonance imaging (MRI)–detectable redox active drug that localizes selectively in tumor cells and enhances the effect of radiation therapy. This phase Ib/II trial of motexafin gadolinium, administered concurrently with 30 Gy in 10 fractions whole-brain radiation therapy (WBRT), was conducted to determine maximum-tolerated dose (MTD), dose-limiting toxicity, pharmacokinetics, and biolocalization in patients with brain metastases. Additional endpoints were radiologic response rate and survival. PATIENTS AND METHODS: Motexafin gadolinium was administered before each radiation treatment in this open-label, multicenter, international trial. In phase Ib, drug dose was escalated until the MTD was exceeded. In phase II, drug was evaluated in a narrow dose range. RESULTS: In phase Ib, the motexafin gadolinium dose was escalated in 39 patients (0.3 mg/kg to 8.4 mg/kg). In phase II, 22 patients received 5 mg/kg to 6.3 mg/kg motexafin gadolinium. Ten once-daily treatments were well tolerated. The MTD was 6.3 mg/kg, with dose-limiting reversible liver toxicity. Motexafin gadolinium’s tumor selectivity was established using MRI. The radiologic response rate was 72% in phase II. Median survival was 4.7 months for all patients, 5.4 months for recursive partitioning analysis (RPA) class 2 patients, and 3.8 months for RPA class 3 patients. One-year actuarial survival for all patients was 25%. CONCLUSION: Motexafin gadolinium was well tolerated at doses up to 6.3 mg/kg, was selectively accumulated in tumors, and, when combined with WBRT of 30 Gy in 10 fractions, was associated with a high radiologic response rate.

1509Background: Genetic testing is a powerful tool for stratifying cancer risk. Multiplex gene panel (MGP) testing allows simultaneous analysis of multiple high- and moderate- penetrance genes. How...

e23162Background: Cholangiocarcinoma is an uncommon gastrointestinal malignancy that remains relatively understudied even as its worldwide incidence continues to climb. The treatment options for pa...

1010Background: A rich lymphocytic infiltrate in pre-treatment tumor biopsies predicts better overall survival (OS) in TNBC. We investigated whether peripheral lymphocyte count is similarly associa...

1503Background: Sequencing more genes increases the chance of finding a pathogenic mutation and/or a variant of uncertain significance (VUS). Little is known about potential harms of multiplex test...

e18036Background: While molecular profiling (MP) is becoming widely accessible, clinician understanding may be a barrier to GMCT enrollment. MP programs at academic medical centers report GMCT enro...

e17576 Background: Nearly 85% of cancer patients (pts) are treated in the community. In order to advance our understanding of cancer genomics and the impact of targeted therapy, this research needs to expand in the community. Community-based MP programs can improve access to targeted therapies and support genomic research. Methods: From June 1st, 2014 through January 15th, 2015 we instituted a prospective MP study at Sanford Health. Sanford Health is a multi-center, predominantly rural health system centered in the Dakotas. Patients with incurable, advanced solid tumors who had progressed after exhausting standard treatment options were offered enrollment in a MP program. All patients enrolled had MP performed on a recent tumor biopsy ( < 14 weeks from enrollment) using the Foundation One targeted sequencing panel. Results were reviewed at a multi-center, video-conferenced, weekly genomic tumor board. Treatment options were offered based on actionable alterations (AA) identified through MP. FDA approved o...

e17521 Background: Precision medicine has the potential to revolutionize cancer care by matching the right patient to the right therapy. Using genomics in routine care to guide therapy selection re...

1018 Background: The HRD-LOH (loss of heterozygosity) score is significantly associated with favorable response to neoadjuvant platinum-based therapy in PrECOG 0105. We set out here to assess the c...

1513 Background: Genetic testing for hereditary breast and ovarian cancer is evolving with the recent introduction of multi-gene panels, although the clinical impact of such panels is as yet poorly...

e17641 Background: The advent of Next-Generation Sequencing (NGS), and other diagnostic technologies, has enabled the use of genomic information to guide targeted treatment in cancer patients. The ...

433 Background: This study investigates which pre-treatment hematologic and tumor marker values can predict for overall survival (OS) in patients with unresectable pancreatic adenocarcinoma treated with stereotactic body radiotherapy (SBRT) using data from the past 11 years from a single institution. Methods: Clinical data from 196 patients with primary unresectable pancreatic adenocarcinoma treated with SBRT at our institution from 2002 to 2013 was collected. Of these, 161 (82.1%) had locally advanced disease, 12 (6.1%) had borderline resectable disease, 21 (10.7%) were medically inoperable, and 2 (1.0%) had resectable disease but refused surgery. Prior to the start of treatment, lab values were collected for 191 of these patients including: hemoglobin, platelets, albumin, red blood cell (RBC) count, white blood cell (WBC) count, absolute lymphocyte count, absolute neutrophil count, and neutrophil to lymphocyte ratio (NLR). Additionally, tumor markers CA 19-9 and CEA were collected. Background clinical and demographic data was also available for the entire cohort. Results: The median follow-up for the entire cohort following SBRT was 7.7 months. Median OS from the start of treatment was 7.88 months (IQR 4.96-12.72). Prior radiation therapy (p = 0.0462), lower RBC (p = 0.0307), higher pre-treatment CEA (p = 0.0166), and higher NL ratio (p < 0.0001) correlated with worse OS on univariate analysis. WBC, platelet counts, and CA 19-9 did not predict for OS on univariate analysis. On multivariate analysis, higher NL ratio (p = 0.0455) and higher CEA (p = 0.0148) were both correlated with decreased OS, while RBC and prior radiation therapy had no association. Conclusions: Review of pre-treatment hematologic and tumor marker values serves as a valuable predictor of overall survival in patients with unresectable pancreatic adenocarcinoma treated with SBRT. Higher pre-treatment NL ratio and CEA values are associated with poorer OS, which may help to stratify patients into different risk categories.

Conflicting reports have created controversy with respect to the effects of a positive family history on the survival of patients who are diagnosed with gastric cancer. Previously, Yatsuya et al reported that a positive family history of gastric cancer was negatively associated with patient survival. Other studies have reported that family history either reduces or has no effect on the risk of death from gastric cancer. In the article that accompanies this editorial, Han et al report on a study that seeks to clarify the relationship between a positive family history and survival of gastric cancer in a large, retrospective review of surgical cases from Korea. The authors retrospectively evaluated 1,273 Korean patients with gastric cancer for the presence of firstor second-degree relatives with any cancer. Results were reported for percentage of patients with first-degree relatives with any cancer (38.6%) or cancer type, including gastric cancer (20.6%). The authors reported a statistically significant improvement in disease-free survival, recurrence-free survival, and overall survival at 5 years for patients with a first-degree family history of gastric cancer compared with those without a positive family history of gastric cancer. These findings are similar to reports in other cancer types (ie, breast and colon) in which a positive family history is associated with improved prognosis. In addition, these results are most consistent with those previously reported by Yatsuya et al wherein a family history of gastric cancer was associated with improved patient survival. This study confirms that a positive family history of gastric cancer is clinically significant and prognostically favorable. These conclusions carry important implications for the families of individuals who are diagnosed with gastric cancer because it affects the mode of screening in first-degree family members, particularly in Asia where individuals with a first-degree family history of gastric cancer undergo more strict screening programs. This study also serves as a reminder of the importance of obtaining a detailed family history, which continues to be something of a lost art. Indeed, a history of cancer in firstor second-degree relatives has important implications for patients and for their family members who would potentially be candidates for more aggressive screening measures, particularly in cases of familial colon, breast, gastric, and other cancers. Despite the new insights about the significance of family history in gastric cancer that are provided by this study, we are left to wonder about the mechanism by which a positive history affects prognosis. Several potential explanations exist, including the possibility that individuals with a family history of gastric cancer share a genetic predisposition for their improved survival. For example, the presence of microsatellite instability (MSI) is a hallmark of all patients with hereditary nonpolyposis colon cancer and is present in 15% of sporadic colon cancers. Furthermore, MSI in colorectal cancer is positively associated with patient survival. MSI also serves as a negative predictive marker for response to treatment in patients with colorectal cancer. The role of MSI has been investigated in gastric cancer and is identified in 13% to 44% of sporadic tumors. Although a trend toward improved survival in MSI-high gastric cancer has been noted in several studies, these observations have not consistently reached statistical significance, and thus it is unclear whether the presence of MSI in gastric tumors serves as a prognostic or predictive indicator as it does in colon cancer. Additional studies are required to definitively determine whether the superior prognosis that has been observed in patients with a positive family history of gastric cancer is attributable to underlying genetic differences such as MSI or other genetic characteristics yet to be identified. The risk of developing gastric cancer is influenced by both genetic and environmental factors. Gastric adenocarcinoma is observed as part of inherited cancer syndromes that include Li-Fraumeni, hereditary nonpolyposis colon cancer, familial adenomatous polyposis, Peutz-Jeghers, and hereditary diffuse gastric cancer. Hereditary diffuse gastric cancer is an inherited gastric cancer predisposition syndrome that is associated with germline transmission of a CDH1 mutation and confers an approximate 75% lifetime risk of developing diffuse-type gastric cancer. Women with CDH1 germline mutations also carry an approximate 40% lifetime risk of developing lobular breast cancer. In contrast, environmental factors such as Helicobacter pylori (H pylori) infection, which confers an approximate five-fold increased risk of developing gastric cancer (primarily intestinal type), significantly affect gastric cancer risk. However, the role of H pylori infection in gastric cancer does not seem to be strictly environmental. Some evidence suggests that genetic factors contribute to an individual’s risk of developing gastric cancer after infection with H pylori. It will be important to determine whether a family history of H pylori infection and gastric cancer influences patient survival in a similar fashion. Additional explanations for the observed differences in prognosis for patients with a family history of gastric adenocarcinoma compared with those without such a history include behavioral differences and variable adherence to screening. A history of tobacco or alcohol use, as noted by the authors, is associated with a poorer prognosis in patients with gastric cancer. The study by Han et al identified a higher rate of smoking cessation and alcoholic consumption in patients with a firstdegree family history, which suggests that this health-related behavioral difference might influence the reported results. In contrast, the rate of adherence to screening recommendations was not observed to be different among those with a first-degree family history of gastric cancer (although this observation did not achieve statistical significance), which suggests that differential adherence to screening cannot account for the observed survival benefit. Editorials

12 Background: Personalized genomic cancer medicine is an approach that has long shown efficacy in molecularly-defined subsets of breast and lung cancers, amongst others, but has not been employed more broadly, in part, due to limitations in testing technologies. Recent advances in genomic technologies have increasingly alleviated these constraints, thereby enabling precision cancer medicine. Data regarding the quality outcomes of patients treated with precision cancer medicine is an important next step along the path to widespread employment of this approach. Methods: We performed an IRB-approved retrospective analysis evaluating the use of targeted therapies matched to patients’ molecular aberrations as determined by genomic testing and recommended by a Molecular Tumor Board (MTB). This study assessed whether a detailed molecular profile of advanced solid malignancies including, but not limited to, the lung, gastrointestinal tract, bladder, prostate, ovary, uterus and skin, that were subsequently treated with matched targeted therapies resulted in improvements in three quality outcome measures: 1) Progression Free Survival (PFS), 2) treatment related morbidity, and 3) cost of treatment. Results: Preliminary results of the cohort analysis suggest that the costs associated with genomic targeted therapy are comparable to standard therapies; however, the costs of treatment related morbidities is significantly lower for patients receiving genomic cancer medicine compared to standard chemotherapy approaches. In line with these findings, overall treatment related morbidities are significantly reduced in the genomic cancer medicine cohort compared to a control cohort. Data regarding the Progression Free Survival are pending at the time of this report. Conclusions: These retrospective data suggest that personalized genomic cancer medicine approaches result in decreased morbidities and cost savings compared to standard chemotherapeutic approaches. In patients with advanced cancer, genomic-based treatments appear to be cost-effective, safe and viable option for treating advanced cancer patients. Additional data regarding survival outcomes are required to determine efficacy of treatment.

1000^ Background: Increased TILs are prognostic and predictive of therapy response in TNBC. PrECOG 0105, a neoadjuvant trial of carboplatin, gemcitabine and iniparib, enrolled 80 pts with clinical ...

115 Background: Gastroesophageal junction (GEJ) and gastric adenocarcinomas constitute a major health problem worldwide. In the United States, the incidence of adenocarcinoma of the esophagus, GEJ and gastric cardia is rising rapidly. Chemotherapy for metastatic disease improves survival and quality of life when compared to best supportive care. The addition of bevacizumab to chemotherapy in a large, randomized, multinational study showed conflicting results. Methods: Patients (pts) with biopsy proven metastatic or unresectable GEJ and gastric adenocarcinomas received bevacizumab 15 mg/kg IV and carboplatin AUC 6 IV on day 1 and capecitabine 850 mg/m2twice daily by mouth on days 1-14 on a 21-day cycle. Pts with stable disease (SD) or partial response (PR) were allowed to continue therapy with bevacizumab with or without capecitabine. Primary endpoints were progression free survival (PFS) and toxicity; secondary endpoints were overall survival (OS) and response rate by RECIST. Results: Between April 2009 and April 2013, a total of 35 pts were enrolled. At present, 29 patients are off treatment and 6 are still actively receiving treatment. 29 pts are available for response assessment, 6 patients came off study prior to their first tumor assessment. Baseline characteristics are: M:F = 25:10; median age = 58 years (range 25-81); primary site = GEJ (17) and gastric (18); degree of differentiation = well (3), moderate (12), poor/signet ring (18), and unknown (2); race = White Non-Hispanic (24), White Hispanic (6), and Asian (5). The 1-year PFS is 21% and median PFS is 8.5 months. Median OS is 14.3 months. Best response by RECIST is as follows: PR in 17 pts (49%), SD in 6 pts (17%), and progressive disease (PD) in 6 pts (17%). The treatment was well-tolerated with toxicities as expected. Conclusions: The combination of capecitabine, carboplatin and bevacizumab in metastatic or unresectable unresectable GEJ and gastric adenocarcinomas is well tolerated and shows promising response rates, though did not meet the primary endpoint of PFS. Clinical trial information: NCT00780494.

7 Background: Multiple-gene sequencing panels are entering clinical practice. We report on research testing with a custom sequencing panel comprising 43 genes and 32 cancer-associated variants among patients referred for assessment of hereditary breast and ovarian cancer risk. Methods: Patients referred to the Stanford Cancer Genetics Program for clinical BRCA1 and BRCA2 (BRCA1/2) mutation testing from 2002-2012 were invited to donate a blood sample for research on an Institutional Review Board-approved protocol. Blood samples were frozen at -80 degrees, and DNA extracted after

1003 Background: TN and BRCA1-deficient breast cancer (BC) cell lines exhibit enhanced sensitivity to DNA damaging agents. This study was designed to assess efficacy, safety and predictors of response to iniparib in combination with GC in early-stage TN and BRCA1/2 mutation-associated BC. Methods: This single-arm, phase II study (NCT00813956) enrolled pts with clinical stage I-IIIA (T ≥ 1cm by MRI) ER-negative (≤ 5%), PR-negative (≤ 5%), and HER2-negative or BRCA1/2 mutation-associated BC. Neoadjuvant G (1000 mg/m2; IV; D1, 8), C (AUC 2; IV; D1, 8), and iniparib (5.6 mg/kg; IV; D1, 4, 8, 11) were given every 21 days for 4 cycles, until the protocol was amended to increase the treatment duration to 6 cycles, with enrollment of 80 pts at multiple PrECOG institutions. The primary endpoint is pathologic complete response (pCR), defined as no invasive carcinoma in the breast and axilla. Pathologic response was centrally assessed by the residual cancer burden (RCB) index. Assuming 76/80 eligible and treated pts, the regimen would be deemed effective if the lower bound of a 90% exact binomial CI on the pCR rate exceeded 25%. Secondary endpoints are safety, MRI response, and breast conservation. Results: Among 80 eligible pts treated with 6 cycles, median age is 48 years, 19 pts have germline BRCA1/2 mutations (90% tested to date) and clinical stage is I (13%), IIA (36%), IIB (36%), IIIA (15%). Pathologic response data (ITT population) are detailed below. 69 pts completed treatment per protocol: 5 progressed, 5 discontinued due to an AE and 1 mutation carrier was lost to follow-up. Most common G3/4 adverse events are neutropenia (49%), elevated ALT/AST (14%), and anemia (10%). Conclusions: Preoperative GC plus iniparib is active in the treatment of early-stage TN and BRCA1/2 mutation-associated BC. Clinical trial information: NCT00813956. [Table: see text]

77 Background: The gastric cancer AJCC staging system recently underwent significant modifications of the T and N categories as well as stage groupings. The new system has not been validated on a US population database, but studies on Asian patients have reported no difference in survival between stages IB and IIA, as well as IIB and IIIA. Methods: California Cancer Registry data linked to Office of Statewide Health Planning and Development discharge abstracts were used to identify patients with gastric adenocarcinoma (gastroesophageal junction tumors excluded) who underwent curative-intent surgical resection from 2002 to 2006. AJCC stage was reclassified based on the 7th edition. Disease-specific survival (DSS) probabilities were calculated using the Kaplan-Meier method and compared using the log rank test. Results: Of 4,985 patients identified, 2,262 had complete pathologic data and known cause of death. Median age was 70 years and 60% were males. Median number of examined lymph nodes was 12 and 39% of patients received adjuvant chemotherapy. The 7th edition AJCC system did not distinguish outcome adequately between stages IB and IIA (P = .25), or IIB and IIIA (P = .33, Table ). By merging stage II into one category and moving T2N1 to stage IB and T2N2, T1N3 to stage IIIA, we propose a new grouping system which showed improved discriminatory ability ( Table ). Conclusions: In this first study validating the new 7th edition AJCC staging system for gastric cancer on a US population, we found stages IB and IIA, as well as IIB and IIIA to perform similarly. We propose a revised stage grouping for the AJCC system that better discriminates between outcomes. [Table: see text]

161 Background: Information from BRCA1/2 mutation testing is used to guide cancer screening and risk reduction strategies. More recently, BRCA1/2 mutation status is being used to select patients for breast cancer (BC) clinical trials employing new targeted therapies, such as PARP inhibitors. There is concern that current testing practices do not produce results within the timeframe necessary for clinical trial eligibility assessment. Methods: We analyzed BRCA1/2 mutation testing patterns in the context of a prospective phase II neoadjuvant study of gemcitabine, carboplatin and BSI-201 (iniparib) for women with Stage I-IIIA triple-negative or BRCA1/2 mutation-associated BC. Specifically, we analyzed: 1) the proportion of participants who met BRCA1/2 mutation testing criteria per NCCN Guidelines v1.2011 and 2) the proportion who had the Myriad Genetics Single Site, Multisite 3 (Ashkenazi Panel), BRACAnalysis and Comprehensive Rearrangement tests performed and their results. For patients without a known BRCA1/2 mutation prior to BC diagnosis who met testing criteria, we assessed the length of time from BC diagnosis to receipt of test results. Results: Sixty-four patients between the ages of 26 and 73 were enrolled. Results are summarized below. For patients who had BRCA1/2 testing after BC diagnosis, the median number of days from BC diagnosis to receipt of results was 42. Conclusions: The frequency of BRCA1/2 mutations in this cohort is high (20%), though nearly half of these mutations were found after patients provided trial consent. Current standard processes for testing women with newly diagnosed BC for BRCA1/2 mutations do not occur expeditiously enough for many to guide clinical trial selection. [Table: see text]

4056 Background: AVAGAST, a large international phase III study, failed to show improved overall survival (OS) with the addition of Bev to chemotherapy for patients (pts) with advanced GC (HR 0.87 ...

1519 Background: Breast cancer (BC) is the most common tumor in women with germ-line TP53 mutations with 49% risk by age 60, and significant risk before age 40. The histopathological profile of BC ...

14 Background: Gastric cancer is the fourth most prevalent malignancy and second leading cause of cancer death worldwide. For all stages, gastric cancer is rapidly fatal with a 5-year survival rate of 25%. Of those patients who survive, very little is known about their demographic and tumor characteristics. Methods: Cases of gastric and gastroesophageal junction (GEJ) adenocarcinoma diagnosed between 1998 and 2005 were identified within the California Cancer Registry, a SEER registry. Univariate and multivariate logistic regression models were constructed predicting the outcome of survival to 3, 5 and 8 years based on demographic, disease and treatment variables. Results: Of the 47,049 patients with gastric or GEJ cancer, 38,380 (82%) died within 3 years of diagnosis, 8,669 (18%) survived at least 3 years, 5,336 (11%) survived at least 5 years, and 3,002 (6%) survived at least 8 years. The 5-year survivors had a median age at diagnosis of 67 years with a male predominance (64%). The majority were non-Hispanic whites (64%), followed by Asian/Pacific Islanders (23%), and Hispanics (18%). Most had localized (48%) or regional (41%) disease. In the multivariate model, odds of surviving to 5 years were highest with localized stage (odds datio=27), surgery (7.8), regional stage (5.9), fundus/body/antrum site (1.7), Asian/Pacific Islander race (1.5), intestinal histology (1.5), high SES (1.4), and well/moderately differentiated tumors (1.4). Kaplan-Meier curves for cause-specific death were examined by anatomic site and histology. Persons with tumors originating in the fundus/body/antrum had longer median OS compared to patients with tumors originating in the esophagus/cardia (15 vs. 12 months). Intestinal histology had longer median OS (30 months) compared to other or diffuse (both approx 12 months). Conclusions: Long-term survivors of gastric and GEJ cancer have distinct demographic patterns when compared to patients who died within three years. The multivariate analyses demonstrated that earlier stage, surgery, anatomic site (fundus/body), Asian race, intestinal histology, high SES, well/moderately differentiated tumors, and radiation were the strongest independent predictors of survival. No significant financial relationships to disclose.

175 Background: Pancreatic cancer is one of the most lethal malignancies, with an estimated 5-year survival rate of 6%, often due to advanced stage at diagnosis. However, there is a small population of patients, even with metastatic disease, who survive 3 years and beyond. Methods: Cases of pancreatic adenocarcinoma diagnosed from 1998 to 2005 were identified in the California Cancer Registry. A multivariate logistic regression model was constructed to predict the outcome of 3+ year survival according to demographic, disease and treatment variables defined a priori: age, gender, race/ethnicity, socioeconomic status (SES), histology, stage, treatment within 4 months of diagnosis (surgery, chemotherapy, and/or radiation), prior cancer history, and treatment at an academic hospital. Results: Among 54,475 cases of pancreatic cancer with ≥3 years of follow-up data available, median survival was 3.5 months. 2,855 patients (5.2%) survived at least 3 years, of whom 19% had remote stage disease at diagnosis. On multivariate analysis, advanced age was associated with decreased odds of long-term survival (p Conclusions: In a well- characterized population-based registry with rigorous follow-up, we were able to identify a cohort of long-term survivors of pancreatic adenocarcinoma as well as factors associated with their exceptional survival. Planned future work with this cohort includes case-control studies making use of tumor and germline specimens, as well as survivorship research. No significant financial relationships to disclose.

4104 Background: Neuroendocrine tumors (NETs) are a rare and heterogeneous class of neoplasms. They frequently metastasize to the liver and have a unique hypervascular appearance, making this class of tumors an interesting target for angiogenesis inhibition. Methods: Patients (pts) with biopsy proven metastatic or unresectable neuroendocrine tumors received bevacizumab 7.5 mg/kg IV and oxaliplatin 130 mg/m2 IV on day 1 and capecitabine 850 mg/m2 twice daily by mouth on days 1-14 on a 21-day cycle. Pts with stable disease (SD) or partial response (PR) were allowed to continue therapy with bevacizumab with or without capecitabine. Primary endpoints were progression free survival (PFS) and toxicity; secondary endpoints were overall survival (OS) and response rate by RECIST. Results: Between December 2006 and October 2009, 40 pts enrolled, 31 pts are available for response assessment, 6 patients came off study prior to their first tumor assessment, and 3 others are not yet evaluable for response. At a median ...

e14590 Background: 42,470 cases of pancreatic cancer are expected in 2009, with a median age of 72 at presentation. Appropriate therapy for elderly patients is not always clear. We report our exper...