Search

Germline testing for highly penetrant mutations in cancer susceptibility genes such as BRCA1 and BRCA2 (BRCA1/2) can prevent cancer and save lives. Inherited BRCA1/2 mutations confer a 39% to 85% lifetime risk of female breast cancer and an 11% to 62% lifetime risk of ovarian cancer. Identifying BRCA1/2 mutation carriers thus allows for prophylactic surgeries, which can markedly decrease cancer incidence, morbidity, and mortality. Despite these benefits and increasing public awareness, a low fraction (35% in a recent Israeli study) of women with high-risk histories have undergone BRCA1/2 testing. Moreover, women from families with few female relatives are unlikely to recognize their risk of carrying a BRCA1/2 mutation until they themselves develop cancer. Recently, citing the impact of riskreducing surgeries, King et al published a high-profile appeal for BRCA1/2 mutation screening to be offered to all US women at age 30 years. King is to be commended for stimulating a serious discussion on the potential impact of bringing modern genetic medicine to the general population in a compelling circumstance. However, because testing all women for BRCA1/2 mutations would represent the first population-based screening for a hereditary cancer syndrome, careful consideration of the potential limitations, risks, and benefits of population-wide BRCA1/2 testing is essential. Currently, genetic screening efforts in the general population examine newborns for rare recessive disorders (eg, phenylketonuria, beta thalassemia) or screen prenatally for carriers of populationdependent recessive conditions (eg, Tay-Sachs, cystic fibrosis), seeking to prevent morbidity and mortality from childhood-onset diseases. In the case of autosomal dominantly inherited BRCA1/2 mutations, the goal is to reduce suffering and mortality from associated adult-onset malignancies, particularly in individuals who would not otherwise be aware of their cancer risk. Demonstration studies in the Ashkenazi Jewish population have shown that screening for its three BRCA1/2 founder mutations (with a combined 1-in-40 prevalence among unselected Ashkenazi Jews) is cost-effective and identifies carriers who would not have been suspected on the basis of their family cancer history. Citing a recent study of population-wide screening for the Ashkenazi founder mutations in Israel, King et al claim that the high penetrance of BRCA1/2 mutations is independent of whether patients who carry the mutations are ascertained on the basis of family history. However, all such studies examined Ashkenazi founder mutations only, which constitute more than 90% of mutations in the Ashkenazi population, thereby limiting generalizability to other populations. In the general US population, in which combined prevalence of a diverse array of BRCA1/2 mutations is 10-fold lower, widespread screening would have different performance characteristics than in the Ashkenazim. Whether the assortment of BRCA1/2 mutations found in non-Ashkenazi populations would have the same penetrance in population-based ascertainment as they do in clinicbased patients is unknown.

Colorectal cancer (CRC) heritability has been estimated to be around 30%. However, mutations in the known CRC susceptibility genes explain CRC risk in under 10% of the cases. Germline mutations in DNA-repair genes (DRGs) have recently been reported in CRC but their contribution to CRC risk is largely unknown. We evaluated the gene-level germline mutation enrichment of 40 DRGs in 680 unselected CRC individuals compared to 27728 ancestry-matched cancer-free adults. Significant findings were then examined in independent cohorts of 1661 unselected CRC cases and 1456 early-onset CRC cases. Of 680 individuals in the discovery set, 31 (4.56%) individuals harbored germline pathogenic mutations in known CRC susceptibility genes while another 33 (4.85%) individuals had DRG mutations that have not been previously associated with CRC risk. Germline pathogenic mutations inATMandPALB2were enriched in both the discovery (OR= 2.81; P= 0.035 and OR= 4.91; P= 0.024, respectively) and validation sets (OR= 2.97; Adjusted P= 0.0013 and OR= 3.42; Adjusted P= 0.034, forATMandPALB2respectively). Biallelic loss ofATMwas evident in all cases with matched tumor profiling. CRC cases also had higher rates of actionable mutations in the HR pathway that can substantially increase the risk of developing cancers other than CRC. Our analysis provides evidence forATMandPALB2as CRC risk genes, underscoring the importance of the homologous recombination pathway in CRC. In addition, we identified frequent complete homologous recombination deficiency in CRC tumors, representing a unique opportunity to explore targeted therapeutic interventions such as PARPi.

Colorectal cancer (CRC) heritability has been estimated to be around 30%. However, mutations in the known CRC-susceptibility genes explain CRC risk in fewer than 10% of affected individuals. Germline mutations in DNA-repair genes (DRGs) have recently been reported in CRC, but their contribution to CRC risk is largely unknown. We evaluated the gene-level germline mutation enrichment of 40 DRGs in 680 unselected CRC individuals and 27,728 ancestry-matched cancer-free adults. Significant findings were then examined in independent cohorts of 1,661 unselected CRC individuals and 1,456 individuals with early-onset CRC. Of the 680 individuals in the discovery set, 31 (4.56%) individuals harbored germline pathogenic mutations in known CRC-susceptibility genes, and another 33 (4.85%) individuals had DRG mutations that have not been previously associated with CRC risk. Germline pathogenic mutations in ATM and PALB2 were enriched in both the discovery (OR = 2.81 and p = 0.035 for ATM and OR = 4.91 and p = 0.024 for PALB2) and validation (OR = 2.97 and adjusted p = 0.0013 for ATM and OR = 3.42 and adjusted p = 0.034 for PALB2) sets. Biallelic loss of ATM was evident in all individuals with matched tumor profiling. CRC individuals also had higher rates of actionable mutations in the HR pathway, which can substantially increase the risk of developing cancers other than CRC. Our analysis provides evidence for ATM and PALB2 as CRC-risk genes, underscoring the importance of the homologous recombination pathway in CRC. In addition, we identified frequent complete homologous recombination deficiency in CRC tumors, representing a unique opportunity to explore targeted therapeutic interventions such as poly-ADP ribose polymerase inhibitor (PARPi).

Cancer development is a complex process driven by inherited and acquired molecular and cellular alterations. Prevention is the holy grail of cancer elimination, but making this a reality will take a fundamental rethinking and deep understanding of premalignant biology. In this Perspective, we propose a national concerted effort to create a Precancer Atlas (PCA), integrating multi-omics and immunity – basic tenets of the neoplastic process. The biology of neoplasia caused by germline mutations has led to paradigm-changing precision prevention efforts, including: tumor testing for mismatch repair (MMR) deficiency in Lynch syndrome establishing a new paradigm, combinatorial chemoprevention efficacy in familial adenomatous polyposis (FAP), signal of benefit from imaging-based early detection research in high-germline risk for pancreatic neoplasia, elucidating early ontogeny in BRCA1-mutation carriers leading to an international breast cancer prevention trial, and insights into the intricate germline-somatic-immunity interaction landscape. Emerging genetic and pharmacologic (metformin) disruption of mitochondrial (mt) respiration increased autophagy to prevent cancer in a Li-Fraumeni mouse model (biology reproduced in clinical pilot) and revealed profound influences of subtle changes in mt DNA background variation on obesity, aging, and cancer risk. The elaborate communication between the immune system and neoplasia includes an increasingly complex cellular microenvironment and dynamic interactions between host genetics, environmental factors, and microbes in shaping the immune response. Cancer vaccines are in early murine and clinical precancer studies, building on the recent successes of immunotherapy and HPV vaccine immune prevention. Molecular monitoring in Barrett's esophagus to avoid overdiagnosis/treatment highlights an important PCA theme. Next generation sequencing (NGS) discovered age-related clonal hematopoiesis of indeterminate potential (CHIP). Ultra-deep NGS reports over the past year have redefined the premalignant landscape remarkably identifying tiny clones in the blood of up to 95% of women in their 50s, suggesting that potentially premalignant clones are ubiquitous. Similar data from eyelid skin and peritoneal and uterine lavage fluid provide unprecedented opportunities to dissect the earliest phases of stem/progenitor clonal (and microenvironment) evolution/diversity with new single-cell and liquid biopsy technologies. Cancer mutational signatures reflect exogenous or endogenous processes imprinted over time in precursors. Accelerating the prevention of cancer will require a large-scale, longitudinal effort, leveraging diverse disciplines (from genetics, biochemistry, and immunology to mathematics, computational biology, and engineering), initiatives, technologies, and models in developing an integrated multi-omics and immunity PCA – an immense national resource to interrogate, target, and intercept events that drive oncogenesis. Cancer Res; 77(7); 1510–41. ©2017 AACR.

693 Background: Paired-related homeodomain transcription factor 1 (Prrx1) isoforms are involved in pancreatic development, pancreatitis, and carcinogenesis. Hepatocyte growth factor (HGF) is a transcriptional target of Prrx1b. Ficlatuzumab is a recombinant humanized HGF antibody, that neutralizes HGF/c-Met binding and HGF-induced c-Met phosphorylation. In preclinical pancreatic ductal adenocarcinoma (PDAC) models, inhibition of Prrx1b-HGF signaling using ficlatuzumab and gemcitabine reduced primary tumor volume and eliminated metastatic disease. Methods: Patients (pts) with previously-untreated metastatic PDAC enrolled in a phase Ib dose escalation study with 3+3 design and two dose cohorts of ficlatuzumab (10mg/kg and 20mg/kg) administered intravenously every other week with gemcitabine (G; 1000mg/m2) and nab-paclitaxel (A; 125mg/m2) given 3 weeks on and 1 week off. This was followed by an expansion phase at the maximally tolerated dose (MTD) of the combination. Results: 24 pts (sex, 12M:12F; median age, 69 years [range, 51-82 years]) were enrolled. No dose-limiting toxicities were identified in the phase 1b (N = 6 pts) and ficlatuzumab at 20mg/kg with GA was advanced to the expansion phase (N = 18 pts). By RECISTv1.1 in the full study population, 7 (29%) pts had partial response, 15 (63%) had stable disease, and 2 (8%) could not be evaluated. Median progression-free survival was 8 months (range, 3-16 months), 4 pts are still on study treatment. The primary toxicities attributed to ficlatuzumab included hypoalbuminemia (grade 3, 21%; any grade, 91%) and edema (grade 3, 8%; any grade, 91%). Nine (38%) of the 24 pts discontinued study treatment due to these toxicities prior to disease progression. Conclusions: The combination of ficlatuzumab with gemcitabine and nab-paclitaxel is associated with durable treatment responses but also significant hypoalbuminemia and edema that may impair treatment tolerability. Serial blood samples were collected for circulating HGF measurements, and mandatory pretreatment biopsies were collected for tumor c-MET pathway activation and 3D organoid culture drug sensitivity testing. Clinical trial information: NCT03316599.

678 Background: MGT identifies cancer susceptibility gene variants in 4-10% of unselected PC pts. Such data have prompted national guidelines to recommend GC and MGT of all PC pts, but the benefits and barriers to implementing systematic testing are unknown. This study’s aim was to study the implementation of universal GC for all PC pts seen in an academic oncology practice. Methods: In 12/2016, all Dana-Farber Cancer Institute (DFCI) gastrointestinal oncologists were recommended to refer all PC pts for GC and MGT. In 10/2018, workflows were changed such that PC patients were automatically scheduled for GC consultation on the same day as their initial oncologic evaluation (unless patients opted out), rather than relying on provider referral. Clinical and germline data were collected on a consecutive cohort of PC pts undergoing GC and MGT from 3/1/2017-3/31/2019. Two additional months (4/1/2019-5/31/2019) were collected for clinical quality assessment purposes. Results: 1305 (48.3/month) PC pts were seen for oncologic new patient visits, 318 (25.1%; 12.1/month) of whom underwent GC. Rates of GC/MGT increased significantly after the 10/2018 workflow change (8.2 PC pts/month [17.2% of all new PC pts seen] versus 20.3 PC pts/month, [40.9% of all new PC pts seen]; p

77 Background: There is an increased use of oral anti-cancer therapies (OACTs) for treatment of gastrointestinal (GI) cancers. While OACTs provide convenience compared to IV agents, they carry similar risks for drug-drug interactions (DDI), toxicities, and unique challenges like adherence and drug access. Patients on OACTs have fewer touch-points with clinicians, requiring more patient ownership of treatment. Pharmacist co-management of pts has been shown to be successful in teaching and monitoring of IV therapy. We sought to assess feasibility of pharmacist co-management for pts prescribed OACTs for treatment of GI cancers. Methods: In 2019, the Dana-Farber GI Cancer Center (GCC) had an embedded pharmacist 8 hrs/week to help with co-management of pts on OACTs. The pharmacist provided (1) in-person and telephone teaching; (2) comprehensive medication reconciliation; (3) DDI review; and (4) supportive care recommendations. Patients were identified by reviewing provider schedules and through provider referrals. The initial teach visit was one-on-one with each patient before initiation, with joint visits with providers thereafter for monitoring and adherence checks. Data were collected to quantify the types of support/recommendation provided by pharmacist and the impact on clinical workflow. Results: After 4 months in the GCC clinic, the pharmacist has co-managed 26 new pts, 61% seen in-person. In initial visits, the pharmacist identified 3 DDI, updated 15 medication lists, and assisted 11 pts/or providers with drug access and drug information. The pharmacist saw 10 of 26 pts for follow up, totaling 21 encounters. The pharmacist assisted in 17 of the 21 encounters with drug access and drug information. Pharmacist spent 20 min/pt on teaching. For follow-up visits, the pharmacist did not additional incur clinic resources as patients were seen with providers. Conclusions: Pharmacist co-management of patients on OACTs is feasible and offers an added safety resource to pts and providers from initial teaching to monitoring. Future research will focus on the impacts of co-management on clinical outcomes, such as the use of emergency/hospital visits, the duration of therapy, and adherence.

102 Background: Combination irinotecan and cetuximab is approved for irinotecan-refractory mCRC; it is unknown if the addition of bevacizumab would improve outcomes. We studied the efficacy and safety of CBI compared with CI in patients (pts) with RAS wildtype, irinotecan-refractory mCRC. Methods: In this multicenter, randomized, double-blind, placebo-controlled phase II trial, pts with RAS wildtype mCRC and no prior anti-epidermal growth factor receptor therapy who failed at least 1 irinotecan-based chemotherapy regimen and received bevacizumab in at least 1 prior line of therapy were randomized 1:1 to irinotecan 180 mg/m2 (or previously tolerated dose), cetuximab 500 mg/m2, and bevacizumab 5 mg/kg vs CI every 2 wks until disease progression, intolerable toxicity, or withdrawal of consent. The primary endpoint was progression free survival (PFS). Multivariable Cox proportional hazard models stratified by number of prior lines of therapy and bevacizumab receipt in immediate prior line were performed. Secondary endpoints included overall survival (OS), objective response rate (ORR), and adverse events (AEs). The study was closed early in January 2018 for reasons related to accrual and funding after enrollment of 36 out of a planned 60 pts. Results: Between July 2015 and December 2017, 36 pts were randomized (19 to CBI, 17 to CI). 34 pts (94%) were treated with 2 or more prior chemotherapy regimens. Baseline characteristics were similar between arms. Median PFS was 9.7 vs 5.5 mo for CBI and CI arms, respectively (log-rank P =0.76; multivariable HR = 0.64; 95% CI, 0.25-1.66). Median OS was 19.7 vs 10.2 mo for CBI and CI (log-rank P= 0.04; multivariable HR = 0.41; 95% CI, 0.15-1.09). ORR was 37% for CBI vs 12% for CI ( P =0.13). Grade 3 or higher AEs occurred in 47% of pts receiving CBI vs 35% for CI ( P =0.46). Conclusions: In this prematurely discontinued trial, there were non-significant increases in PFS and ORR and a statistically significant 9.5 mo increase in median OS in favor of CBI compared to CI. Further investigation of CBI for treatment of irinotecan-refractory mCRC is warranted. Clinical trial information: NCT02292758.

TPS4162 Background: 4-10% of PDAC patients harbor pathogenic germline variants in cancer susceptibility genes, including APC, ATM, BRCA1, BRCA2, CDKN2A, EPCAM, MLH1, MSH2, MSH6, PALB2, PMS2, STK11, and TP53. For families with such pathogenic variants, the greatest potential impact of germline testing is to identify relatives with the same pathogenic variant (cascade testing), thereby providing the opportunity for early detection and cancer interception of PDAC and other associated malignancies. Numerous factors limit cascade testing in real-world practice, including family dynamics, widespread geographic distribution of relatives, access to genetic services, and misconceptions about the importance of germline testing, such that the preventive benefits of cascade testing are often not fully realized. The primary aim of this study is to analyze two alternative strategies for cascade testing in families with inherited PDAC susceptibility. Methods: 1000 individuals (from approximately 200 families) with a confirmed pathogenic germline variant in any of the above genes in a 1st/2nd degree relative and a 1st/2nd degree relative with PDAC will be remotely enrolled through the study website (www.generatestudy.org) and randomized between two different methods of cascade testing (individuals with prior genetic testing will be ineligible): Arm 1 will undergo pre-test genetic education with a pre-recorded video and live interactive session with a genetic counselor via a web-based telemedicine platform (Doxy.me), followed by germline testing through Color Genomics; Arm 2 will undergo germline testing through Color Genomics without dedicated pre-test genetic education. Color Genomics will disclose results to study personnel and directly to participants in both arms. Participants in both arms will have the option of pursuing additional telephone-based genetic counseling through Color Genomics. The primary outcome will be uptake of cascade testing. Secondary outcomes will include participant self-reported genetic knowledge, cancer worry, distress, decisional preparedness, familial communication, and screening uptake, which will be measured via longitudinal surveys. Enrollment will begin February, 2019. Clinical trial information: NCT03762590.

Background: Current data suggest that platinum-based combination therapy is the standard first-line treatment for biliary tract cancer. EGFR inhibition has proven beneficial across a number of gastrointestinal malignancies; and has shown specific advantages among KRAS wild-type genetic subtypes of colon cancer. We report the combination of panitumumab with gemcitabine (GEM) and oxaliplatin (OX) as first-line therapy for KRAS wild-type biliary tract cancer. Methods: Patients with histologically confirmed, previously untreated, unresectable or metastatic KRAS wild-type biliary tract or gallbladder adenocarcinoma with ECOG performance status 0–2 were treated with panitumumab 6 mg kg−1, GEM 1000 mg m−2 (10 mg m−2 min−1) and OX 85 mg m−2 on days 1 and 15 of each 28-day cycle. The primary objective was to determine the objective response rate by RECIST criteria v.1.1. Secondary objectives were to evaluate toxicity, progression-free survival (PFS), and overall survival. Results: Thirty-one patients received at least one cycle of treatment across three institutions, 28 had measurable disease. Response rate was 45% and disease control rate was 90%. Median PFS was 10.6 months (95% CI 5–24 months) and median overall survival 20.3 months (95% CI 9–25 months). The most common grade 3/4 adverse events were anaemia 26%, leukopenia 23%, fatigue 23%, neuropathy 16% and rash 10%. Conclusions: The combination of gemcitabine, oxaliplatin and panitumumab in KRAS wild type metastatic biliary tract cancer showed encouraging efficacy, additional efforts of genetic stratification and targeted therapy is warranted in biliary tract cancer.

Survivors of childhood cancers are at increased risk of developing secondary gastrointestinal cancers, including colorectal cancer, later in life, possibly from exposure to abdominopelvic radiotherapy and/or alkylating chemotherapy. Profuse gastrointestinal polyposis is associated with rare, inherited colorectal cancer predisposition syndromes, most commonly caused by mutations in the adenomatous polyposis coli (APC) or mutY homolog (MUTYH) genes. We describe 5 patients who developed gastrointestinal polyposis many years after radiotherapy and chemotherapy for a childhood cancer. Genetic analysis of all 5 subjects did not identify pathogenic germline mutations in APC or MUTYH. Chemotherapy and/or radiotherapy therefore might cause gastrointestinal polyposis in some patients by undiscovered mechanisms.

The original version of this article unfortunately contained a mistake. The name of "Matthew Yurgelun" is now corrected in the author group of this article.

Study of the biology of tumors caused by germline mutations has led to recent paradigm-changing therapy and is driving precision prevention efforts, including immune oncology and early detection research. Here, we explore recent biologic advances that are redefining the spectrum of cancers linked to various hereditary predisposition syndromes and can be leveraged to improve personalized risk assessment and develop novel interventions to prevent or intercept cancer.

Colorectal cancer (CRC) is the third most common cancer diagnosed in men and women and approximately 5% of cases are associated with identifiable germline mutations associated with hereditary cancer syndromes. Lifetime risks for CRC can approach 50%-80% for mutation carriers in the absence of endoscopic and/or surgical intervention, and early identification of at-risk individuals can guide clinical interventions for cancer prevention and treatment. Personal and family history and molecular phenotype of CRC tumors are used in determining which patients should be referred for clinical genetic evaluation. Outcomes of genetic testing performed using next-generation sequencing (NGS) multigene panels suggest there can be significant overlap in clinical features among the various hereditary cancer syndromes. This review summarizes new developments in diagnosis and management of patients with genetic predisposition to CRC.

1512Background: MGT will identify cancer susceptibility gene variants in 4-10% of unselected PC pts. Such data have prompted calls for universal GC and MGT of all PC pts, but the real-world benefit...

e16072Background: There are limited treatment (tx) options for metastatic esophageal cancer patients (pts). We sought to determine the safety and efficacy of pembro among US pts with previously tre...

1517Background: LS is the one of the most common inherited causes of cancer and predisposes to a wide variety of cancers. UTCs (kidney, renal pelvis, ureter, and bladder cancer) are collectively th...

270 Background: Sorafenib is the standard first line treatment for advanced HCC and showed a median time to progression (TTP) of 5.5 months and an overall response rate (ORR) of 2% in the phase III SHARP trial. FOLFOX has shown modest activity in HCC with a progression free survival (PFS) of 2.9 months and ORR of 8% in a phase III trial. In this single-arm, multicenter phase 2 and biomarker study, sorafenib plus FOLFOX was evaluated in the first line treatment of advanced HCC. Methods: Patients with histologically proven advanced HCC, Child Pugh A liver function, and no prior systemic therapies received sorafenib 400mg orally twice daily during a 2-week lead-in, followed by concurrent modified FOLFOX (5-FUCI 1200mg/m2/day for 46 hours and LV 400mg/m2 bolus, Oxaliplatin 85mg/m2) on day 1 and 15 of each 28-day cycle. The primary endpoint was TTP, calculated from date of study entry to date of radiological or clinical disease progression. Serial plasma anti-angiogenic and anti-inflammatory biomarkers were evaluated. Results: The study enrolled 40 patients with advanced HCC: median age, 65 years; male 85%; Child Pugh A5, 70%; BCLC stage C, 95%; HCC etiology, HCV 40%, HBV 13%, alcohol 13%. Grade 3/4 adverse events were notable for AST (23%), ALT (15%), bilirubin (10%), diarrhea (10%), anemia (10%), hypertension (5%), hand-foot syndrome (5%), and thrombocytopenia (5%). Dose reductions for sorafenib and FOLFOX were done in 73% and 65% of patients, respectively. The median TTP was 8.8 months (95%CI, 6.5-11.2). The ORR was 18%, and the stable disease rate was 55%. Among 36 patients with a baseline AFP ≥ 5 ng/mL, 10 (28%) had a ≥ 50% drop in AFP. Low baseline plasma levels of sVEGFR1, VEGF-C, and bFGF and high levels of s-cMET and IL-12 tended to associate with longer TTP (p < 0.10). Decreased s-cMET at day 15 and decreased s-cMET and IL-2 at day 43 were associated with longer TTP (p < 0.05). Conclusions: Sorafenib+FOLFOX demonstrated encouraging clinical efficacy with moderate toxicity in the first line treatment of advanced HCC. Initial biomarker evaluation suggested a correlation between TTP and baseline angiogenic markers as well as changes in IL-2 and s-cMET. Complete biomarker analysis will be presented at the meeting. Clinical trial information: NCT01775501.

Background & Aims Multigene panels are commercially available tools for hereditary cancer risk assessment that allow for next-generation sequencing of numerous genes in parallel. However, it is not clear if these panels offer advantages over traditional genetic testing. We investigated the number of cancer predisposition gene mutations identified by parallel sequencing in individuals with suspected Lynch syndrome. Methods We performed germline analysis with a 25-gene, next-generation sequencing panel using DNA from 1260 individuals who underwent clinical genetic testing for Lynch syndrome from 2012 through 2013. All patients had a history of Lynch syndrome–associated cancer and/or polyps. We classified all identified germline alterations for pathogenicity and calculated the frequencies of pathogenic mutations and variants of uncertain clinical significance (VUS). We also analyzed data on patients' personal and family history of cancer, including fulfillment of clinical guidelines for genetic testing. Results Of the 1260 patients, 1112 met National Comprehensive Cancer Network (NCCN) criteria for Lynch syndrome testing (88%; 95% confidence interval [CI], 86%–90%). Multigene panel testing identified 114 probands with Lynch syndrome mutations (9.0%; 95% CI, 7.6%−10.8%) and 71 with mutations in other cancer predisposition genes (5.6%; 95% CI, 4.4%−7.1%). Fifteen individuals had mutations in BRCA1 or BRCA2 ; 93% of these met the NCCN criteria for Lynch syndrome testing and 33% met NCCN criteria for BRCA1 and BRCA2 analysis ( P = .0017). An additional 9 individuals carried mutations in other genes linked to high lifetime risks of cancer (5 had mutations in APC , 3 had bi-allelic mutations in MUTYH , and 1 had a mutation in STK11 ); all of these patients met NCCN criteria for Lynch syndrome testing. A total of 479 individuals had 1 or more VUS (38%; 95% CI, 35%–41%). Conclusions In individuals with suspected Lynch syndrome, multigene panel testing identified high-penetrance mutations in cancer predisposition genes, many of which were unexpected based on patients' histories. Parallel sequencing also detected a high number of potentially uninformative germline findings, including VUS.

Commercial laboratories now offer an array of multigene panels for hereditary cancer assessment. The authors of this commentary discuss how genetic counseling for inherited cancer risk must rapidly...

3506 Background: In prospective observational studies of mCRC patients, higher plasma levels of 25-hydroxyvitamin D have been associated with improved progression-free (PFS) and overall survival (OS), but the role of vitamin D supplementation in the treatment of mCRC is unknown. Methods: SUNSHINE was a multi-center double-blind phase II randomized controlled trial in previously untreated mCRC patients. Patients were eligible if they had histologically confirmed mCRC, no prior therapy for metastatic disease, ECOG PS 0-1, and were not taking vitamin D >2,000 IU/day x 1 year. All subjects received standard treatment with mFOLFOX6 + bevacizumab with 1:1 randomization to concurrent: HiVitD (vitamin D3 po 8,000 IU/d x 2 wks as loading dose followed by 4,000 IU/d) or LowVitD (standard vitamin D3 400 IU/d) until disease progression, intolerable toxicity, or withdrawal of consent. The primary endpoint was PFS, with the sample size designed to provide 80% power to detect a HR of 0.66 for PFS at a 1-sided alpha=0.2. Results: From April 2012 to November 2016, 139 patients were randomized. Median age was 54 yrs (range 24-82), 57% were male, 77% were white, and 7% had received prior adjuvant chemo. Baseline characteristics were balanced between arms except ECOG PS = 0 was 42% vs. 60% in HiVitD vs. LowVitD. Median follow-up was 16.1 mos (range 0-45.9) and median compliance with VitD capsules was 98%. Patients randomized to HiVitD experienced longer PFS than those receiving LowVitD (median PFS, 12.4 vs. 10.7 mos, respectively; log rank P=0.03). After multivariate adjustment for prognostic variables, HR was 0.66 (95% CI, 0.45-0.99, 2-sided P=0.04). Comparing HiVitD vs LowVitD, RR was 58% vs. 63% ( P=0.54) and disease control rate was 100% vs. 94% ( P=0.05). The most common grade 3-4 toxicities were as expected for FOLFOX-bevacizumab, and none were related to vitamin D. Currently, 14 patients are still actively receiving treatment, and OS data are not yet mature. Conclusion: SUNSHINE met its prespecified primary endpoint, with patients randomized to HiVitD experiencing longer PFS compared to those randomized to LowVitD. A larger confirmatory phase III randomized trial appears warranted. Clinical trial information: NCT01516216.

1513 Background: New insights into lung cancer genomics has required a new look at genetic determinants of inherited lung cancer risk. There are now four germline mutations in lung cancer oncogenes (EGFR T790M, V843I, R776H; and ERBB2 G660D) which have been reported in kindreds with a high prevalence of lung cancer in nonsmokers. We hypothesized that routine tumor NGS could be used to find rare kindreds with inherited lung cancer risk. Methods: An institute-wide database of tumor NGS results was queried for cases positive for one of the reported germline risk alleles in EGFR or ERBB2. At our center, tumor NGS is performed using a hybrid-capture platform spanning exons and key introns of ~400 cancer related genes, and is available for patients (pts) with consent to a research protocol. For living pts with a mutation detected at an allelic fraction (AF) > 25%, and excluding acquired T790M, genetic counseling and CLIA germline testing was provided. For pts with EGFR T790M, testing was performed on the INHERIT EGFR study (NCT01754025; ALCMI). Results: 51 cases were identified from a total of 13,488 cancers with NGS results: 45 with EGFR T790M (all lung cancer), 4 with EGFR R776H (NSCLC, endometrial cancer, and 2 glioma), 2 with ERBB2 G660D (NSCLC and glioblastoma). 34 pts with T790M detected after EGFR inhibitor were excluded, leaving 17 pts of interest. Germline testing was performed on 9 living pts with > 25% AF on NGS, and was positive in 8; 6 of these had germline EGFR T790M. Additionally, germline EGFR R776H was found in a never-smoker with metastatic endometrial cancer; she had a prior a history of NSCLC, and family history was notable for multiple members with lung, breast, and colon cancer. Germline ERBB2 G660D was found in a young never-smoker with metastatic NSCLC and a family history significant for lung cancer in multiple first-degree relatives. Conclusions: Rare germline lung cancer risk mutations in EGFR and ERBB2 can be identified on routine tumor NGS, and may indicate a risk of inherited lung cancer. Our study includes the second known report of a germline EGFR R776H or ERBB2 G660D mutations. Commercial germline NGS assays could be expanded to cover these rare but potentially high-penetrance variants.

Background: Germline genetic testing for hereditary cancer predisposition syndromes can prevent cancer and save lives by identifying individuals who will benefit from enhanced surveillance, prophylactic surgery, and other risk-reduction strategies. Classically, hereditary cancer risk assessment has involved identification of individuals with clinical histories that fulfill specific criteria for a given syndrome followed by germline testing for mutations in the associated gene(s). Recent advances in next-generation sequencing technologies now allow individuals to undergo germline evaluation of numerous genes in parallel and a growing array of multigene panels are clinically available for such diagnostic testing. Methods: This presentation will discuss updates in clinical genetic testing, specifically the growing availability of next-generation sequencing multigene panel testing for hereditary cancer risk assessment. Findings: Data regarding next-generation sequencing multigene panel testing will be presented and will be contrasted with traditional, syndrome-specific hereditary colorectal cancer risk assessment strategies. Particular areas of focus in this talk will include: the detection of unexpected/off-target germline mutations by multigene panels (e.g. BRCA1/2 mutations in patients with colorectal cancer); cost-effectiveness considerations; the significance of moderate-penetrance cancer susceptibility gene mutations; germline variants of uncertain significance; and ways in which multigene panel testing results are expanding knowledge (and uncertainty) about both well- and poorly-defined colorectal cancer predisposition syndromes. Conclusions: Next-generation sequencing multigene panels are rapidly growing in popularity and availability for assessing hereditary cancer risk, including the identification of germline mutations linked to inherited colorectal cancer risk. In order for patients considering such testing to receive optimal care, it is essential that both they and their providers have a clear understanding of the benefits and potential limitations of multigene panels, including the possibility of identifying unexpected and/or indeterminate germline results. Citation Format: Matthew B. Yurgelun. Germline next-generation sequencing panels for detection of inherited colorectal cancer syndromes. [abstract]. In: Proceedings of the AACR Special Conference on Colorectal Cancer: From Initiation to Outcomes; 2016 Sep 17-20; Tampa, FL. Philadelphia (PA): AACR; Cancer Res 2017;77(3 Suppl):Abstract nr IA07.

1501Background: Microsatellite instability (MSI) and mismatch repair protein immunohistochemistry (MMR IHC) are standard of care to screen for the 3% of CRCs that arise due to Lynch syndrome (LS). ...

1521Background: Identifying LS in pts without a prior history of an associated cancer (unaffected pts) relies exclusively on family history analysis and is thus inherently different from LS evaluat...

Objective Due to the increased lifetime risk of endometrial cancer (EC), guidelines recommend that women with Lynch syndrome (LS) age≥35 undergo annual EC surveillance or prophylactic hysterectomy (PH). The aim of this study was to examine the uptake of these risk-reducing strategies. Methods The study population included women meeting clinical criteria for genetic evaluation for LS. Data on cancer risk-reducing behaviors were collected from subjects enrolled in two distinct studies: (1) a multicenter cross-sectional study involving completion of a one-time questionnaire, or (2) a single-center longitudinal study in which subjects completed questionnaires before and after undergoing genetic testing. The main outcome was uptake of EC risk-reducing practices. Results In the cross-sectional cohort, 58/77 (75%) women at risk for LS-associated EC reported engaging in EC risk-reduction. Personal history of genetic testing was associated with uptake of EC surveillance or PH (OR 17.1; 95% CI 4.1–70.9). Prior to genetic testing for LS, 26/40 (65%) women in the longitudinal cohort reported engaging in EC risk-reduction. At one-year follow-up, 16/16 (100%) mismatch repair (MMR) gene mutation carriers were adherent to guidelines for EC risk-reduction, 9 (56%) of whom had undergone PH. By three-year follow-up, 11/16 (69%) MMR mutation carriers had undergone PH. Among women with negative or uninformative genetic test results, none underwent PH after testing. Conclusions Genetic testing for LS is strongly associated with uptake of EC risk-reducing practices. Women found to have LS in this study underwent prophylactic gynecologic surgery at rates comparable to those published for BRCA1/2 mutation carriers.

Colorectal cancers associated with Lynch syndrome are characterized by deficient DNA mismatch repair (MMR) function. Our aim was to evaluate the prevalence of microsatellite instability (MSI) and loss of MMR protein expression in Lynch syndrome–associated polyps. Sixty-two colorectal polyps—37 adenomatous polyps, 23 hyperplastic polyps, and 2 sessile serrated polyps (SSP)—from 34 subjects with germline MMR gene mutations were tested for MSI using a single pentaplex PCR for five mononucleotide repeat microsatellite markers, and also for expression of MLH1, MSH2, MSH6, and PMS2 proteins by immunohistochemistry. High-level MSI (MSI-H) was seen in 15 of 37 (41%) adenomatous polyps, one of 23 (4%) hyperplastic polyps, and one of two (50%) SSPs. Loss of MMR protein expression was seen in 18 of 36 (50%) adenomatous polyps, zero of 21 hyperplastic polyps, and zero of two SSPs. Adenomatous polyps 8 mm or larger in size were significantly more likely to show MSI-H [OR, 9.98; 95% confidence interval (CI), 1.52–65.65; P = 0.02] and deficient MMR protein expression (OR, 3.17; 95% CI, 1.20–8.37; P = 0.02) compared with those less than 8 mm in size. All (six of six) adenomatous polyps 10 mm or larger in size showed both MSI-H and loss of MMR protein expression by immunohistochemistry. Our finding that the prevalence of MMR deficiency increases with the size of adenomatous polyps suggests that loss of MMR function is a late event in Lynch syndrome–associated colorectal neoplasia. Although testing large adenomatous polyps may be of value in the diagnostic evaluation of patients with suspected Lynch syndrome, the absence of an MMR-deficient phenotype in an adenoma cannot be considered as a strong evidence against Lynch syndrome, as it is with colorectal carcinomas. Cancer Prev Res; 5(4); 574–82. ©2012 AACR.

4 Background: FOLFOX has RR 53% and PFS 6.8 months in MEGA (CALGB 80403). VEGF inhibition has improved survival in some but not all randomized trials in MEGA. Ziv-aflibercept binds VEGF-A, B and PlGF and has improved survival in refractory metastatic colorectal cancer. Methods: All pts received mFOLFOX6 q14d. Pts were randomized 2:1 to A) ziv-aflibercept 4mg/kg/d1 or B) placebo. Pts were restaged every 8wks. Primary endpoint: 6mos PFS. Results: 64 pts enrolled (Jan 2013-Apr 2015): sex: 55M/9F; age (median): 32-83 (62); ECOG PS 0/1/2: 33/28/3; primary tumor: esophagus 26/GEJ 18/gastric 20; measurable/evaluable: 52/12; metastases: LN 48/liver 25/lung 15, other 12/adrenal 4/bone 3. Of 64 pts: 34 POD, 9 off for tox, 4 died on tx, 3 withdrew, 4 other, 10 on tx (see Table). Conclusion: Ziv-aflibercept did not significantly improve the efficacy of FOLFOX. HTN was predictive of response to ziv-aflibercept. Except for HTN, there was no significant difference in toxicity between tx arms. Clinical trial information: NCT01747551. [Table: see text]

4038 Background: In MEGCA, cisplatin+CAP has RR 46% and OS 12.1 mos with BEV (AVAGAST) and RR 47% and OS 13.8 mos with TRAS (ToGA). We sought to improve these results by combining CAPOX with both B...

e15124 Background: Advanced CCA carries a poor prognosis, and no standard treatment exists beyond the first line gemcitabine/platinum-based regimen. We performed a single arm phase II and biomarker...

Importance Li-Fraumeni syndrome, usually characterized by germline TP53 mutations, is associated with markedly elevated lifetime risks of multiple cancers, and has been linked to an increased risk of early-onset colorectal cancer. Objective To examine the frequency of germline TP53 alterations in patients with early-onset colorectal cancer. Design, Setting, and Participants This was a multicenter cross-sectional cohort study of individuals recruited to the Colon Cancer Family Registry (CCFR) from 1998 through 2007 (genetic testing data updated as of January 2015). Both population-based and clinic-based patients in the United States, Canada, Australia, and New Zealand were recruited to the CCFR. Demographic information, clinical history, and family history data were obtained at enrollment. Biospecimens were collected from consenting probands and families, including microsatellite instability and DNA mismatch repair immunohistochemistry results. A total of a 510 individuals diagnosed as having colorectal cancer at age 40 years or younger and lacking a known hereditary cancer syndrome were identified from the CCFR as being potentially eligible. Fifty-three participants were excluded owing to subsequent identification of germline mutations in DNA mismatch repair genes (n = 47) or biallelic MUTYH mutations (n = 6). Interventions Germline sequencing of the TP53 gene was performed. Identified TP53 alterations were assessed for pathogenicity using literature and international mutation database searches and in silico prediction models. Main Outcomes and Measures Frequency of nonsynonymous germline TP53 alterations. Results Among 457 eligible participants (314, population-based; 143, clinic-based; median age at diagnosis, 36 years [range, 15-40 years]), 6 (1.3%; 95% CI, 0.5%-2.8%) carried germline missense TP53 alterations, none of whom met clinical criteria for Li-Fraumeni syndrome. Four of the identified TP53 alterations have been previously described in the literature in probands with clinical features of Li-Fraumeni syndrome, and 2 were novel alterations. Conclusions and Relevance In a large cohort of patients with early-onset colorectal cancer, germline TP53 mutations were detected at a frequency comparable with the published prevalence of germline APC mutations in colorectal cancer. With the increasing use of multigene next-generation sequencing panels in hereditary cancer risk assessment, clinicians will be faced with the challenge of interpreting the biologic and clinical significance of germline TP53 mutations in families whose phenotypes are atypical for Li-Fraumeni syndrome.

Giant cell tumor (GCT) of bone is a unique bone lesion that is characterized by an excessive number of multinucleated osteoclasts. GCT consists of neoplastic stromal cells, multinucleated osteoclasts and their precursors, thus serving as a naturally occurring human disease model for the study of osteoclastogenesis. It still remains unclear how stromal cells of GCT recruit osteoclast precursors. In the present study, we characterized the cellular components of GCT and confirmed the presence of CD14 + -monocytes/CD68 + -macrophages and CD34 + -hematopoetic stem cells that express CXCR4, a specific receptor for SDF-1; SDF-1 gene expression and presence of SDF-1 protein were confirmed by real time RT-PCR, in situ hybridization, and immunohistochemistry in the GCT tissue and cultured cells. SDF-1 was present at 25–50 ng/ml in the conditioned media from the GCT cultures, which is in the range of physiological chemotactic concentration. Migration of osteoclast precursors was 2.5-fold higher in response to GCT conditioned media compared to the control media; and migration was inhibited by an average of 36% with anti-SDF-1 neutralizing antibody or competing recombinant SDF-1. These results suggest that SDF-1 is one of the significant chemoattractant factors involved in the recruitment of hematopoietic osteoclast precursor cells during tumor-induced osteoclastogenesis.

800 Background: Cholangiocarcinoma (CCA) carries a poor prognosis in advanced stages, and no standard treatment exists beyond the first line gemcitabine/platinum based combination regimen. Based on preclinical evidence of dysregulation of the VEGF and c-MET pathways in CCA, we performed a single arm phase II study of the dual VEGF receptor and c-MET small molecule inhibitor cabozantinib in patients with advanced CCA. Methods: All eligible pts had histologically confirmed unresectable or metastatic CCA and had progressed after 1 or 2 lines of systemic therapy. Pts were treated with cabozantinib 60mg orally daily continuously for 28 day cycles. Tumor assessments were obtained radiologically every 8 weeks. The primary endpoint was progression free survival (PFS). A Simon two stage design was used, and 9 of 20 pts had to be progression free at 16 weeks in order to proceed to the second stage. Results: Nineteen pts (male 32%; median age 67 years old; intrahepatic vs extrahepatic CCA, 95 vs 5%; 1 vs 2 previous lines of systemic therapy, 53% vs 47%%) were enrolled and all received at least one dose of study drug. After 12 pts failed to be progression free at 16 weeks, the study was terminated as it was determined that the criterion for proceeding to stage 2 could not be met. The median follow-up was 4.07 months (mos). No objective responses were seen. The median PFS was 1.77 mos (95% CI, 1.63-5.37), and the median overall survival was 5.20 mos (95% CI, 2.70-8.17). Eleven (58%) patients required dose reductions. Grade 3 and 4 adverse events occurred in 79% of pts and included neutropenia, epistaxis, hyperbilirubinemia, alkaline phosphatemia, AST/ALT elevation, anemia, lipasemia, and hypertension. One pt had a bowel perforation and another developed an enterocutaneous fistula while on the study drug. VEGF and c-MET pathway-related biomarker analysis is ongoing. Conclusions: Cabozantinib demonstrated limited antitumor activity in an unselected population of pts with advanced CCA. Despite close monitoring, a high rate of > grade 3 toxicities was observed at 60 mg daily dosing. This pt population may be more sensitive to drug doses tolerated by other solid tumor patient populations, and therefore, close monitoring remains critical for these pts in early phase clinical trials. Clinical trial information: NCT01954745.

1509 Background: Multigene panels are increasingly used for assessing hereditary cancer risk due to their ability to analyze numerous cancer susceptibility genes in parallel. Our aim was to study t...

255 Background: Biliary Tract Cancer (BTC) encompasses a group of aggressive, genetically heterogeneous tumors with limited systemic treatment options. Currently platinum and gemcitabine-based therapy is the standard first-line treatment. EGFR inhibition has proven beneficial across a number of gastrointestinal malignancies; and has shown specific advantages among KRASwild-type genetic subtypes of colon cancer. Methods: Patients with histologically confirmed, previously untreated unresectable or metastatic KRAS wild type biliary tract or gallbladder adenocarcinoma with ECOG performance status (PS) 0-2 were eligible. Patients were treated with Panitumumab 6mg/kg, Gemcitabine (GEM) 1000 mg/m2 (10 mg/m2/min), Oxaliplatin (OX) 85 mg/m2on days 1 and 15 of each 28 day cycle. The primary objective was to determine the objective response rate (ORR) by RECIST criteria v.1.1. Secondary objectives were to evaluate toxicity, progression free survival (PFS), and overall survival. Results: Of the 38 patients screened, 31 patients were found to have KRAS wild type genotype and enrolled. The median age was 61 years old, 55% males, 100% of patients had an ECOG PS of

Background Lynch syndrome is characterized by germline mutations in DNA mismatch repair (MMR) genes and carries up to a 70% lifetime risk of colorectal cancer. Impaired MMR gene function results in an abundance of small aberrant nucleotide repeat sequences termed microsatellite instability (MSI). MSI is present in 80-85% of colorectal cancers associated with Lynch syndrome. Prior studies have demonstrated that MMR gene function (as measured by immunohistochemistry) is often lost in Lynch-associated adenomas but is preserved in hyperplastic polyps. The aim of our study was to evaluate the prevalence of MSI in adenomatous and hyperplastic polyps from individuals with Lynch Syndrome.

Background Women with hereditary breast and ovarian cancer (HBOC) have an estimated 15-65% lifetime risk of ovarian cancer; similarly, women with Lynch syndrome have a 40-60% lifetime risk of endometrial cancer and a 10-12% lifetime risk of ovarian cancer. The aim of this study was to investigate the impact of genetic testing on risk-reducing behavior for gynecologic malignancies in women being tested for HBOC and Lynch syndrome.