Clinical application of a functional 3D ex vivo test to predict therapeutic response in patients with HGG: A progression-free survival analysis

2031

Background: High grade gliomas (HGG) including glioblastoma (GBM) are among the most aggressive brain cancers, with patients exhibiting highly variable treatment responses in both newly diagnosed (ND) and recurrent disease. Temozolomide (TMZ) + radiation therapy is the guideline directed standard of care (SOC) in the ND setting; it has remained relatively unchanged for > 15 years, despite variable patient responses. Current available biomarkers do not inform personalized therapy. Functional drug response testing using patient specific tumor cells may have the potential to inform effective therapy selection, thus advancing functional precision oncology. Progression free survival (PFS) is a meaningful surrogate to overall survival (OS) in GBM and therefore, represents a measure of clinical benefit. Methods: The 3D-PREDICT clinical study (NCT03561207) allows enrollment of HGG patients with ECOG ≤ 3, perhaps representing a more accurate real-world population compared to most clinical studies in the same patient cohort. Tumor tissue was prospectively collected during SOC biopsy/resection and analyzed in an ex vivo cell culture test using a panel of agents prescribed as HGG therapeutics. Results: Data pertain to 56 3D-PREDICT patients who had > 6 months of follow up as of December 31, 2021 or experienced progression/death < 6 months post tissue collection. There were 42 3D-PREDICT patients who had IDH wild type ND GBM and received SOC. PFS analysis of these patients showed 3D ex vivo testing was able to prospectively predict TMZ clinical responders vs. non-responders (Kaplan-Meier, p = 0.039; HR 0.516, 95% CI 0.234,1.137). Test predicted TMZ responders had a relative PFS advantage of 3.7 months. Of the 42 patients, 38 had known MGMT methylation status with 23 patients (60%) being unmethylated. Test predicted responders (9) included unmethylated patients; test predicted non-responders (33) included methylated patients. The data suggest ex vivo testing of patient specific tumor tissue may identify ND HGG patients a priori who respond to TMZ, irrespective of MGMT methylation status. Beyond TMZ response prediction, the test assesses tissue response to 11 additional known HGG therapies and therefore may provide a tool to inform potential alternative treatments to TMZ for ND patients. Also evaluated were 14 3D-PREDICT recurrent HGG patients who received test directed salvage therapy; mean PFS after tissue collection from re-resection was 9.0 months (range 2.3 – 24.7 mo) for 9 patients at first recurrence, and 5.7 months (range 1.8 – 11.4 mo) for 5 patients with 2 to 6 recurrences. These examples demonstrate initial test utilization informing salvage therapy selection, correlated with PFS improvements in recurrent HGG. Conclusions: This functional 3D ex vivo cell culture platform provided survival benefit in analyzed ND and recurrent cohorts. Clinical trial information: NCT03561207.

A Functional Precision Medicine Pipeline Combines Comparative Transcriptomics and Tumor Organoid Modeling to Identify Bespoke Treatment Strategies for Glioblastoma

Li Fraumeni syndrome (LFS) is a hereditary cancer predisposition syndrome caused by germline mutations in TP53. TP53 is the most common mutated gene in human cancer, occurring in 30-50% of glioblastomas (GBM). Here, we highlight a precision medicine platform to identify potential targets for a GBM patient with LFS. We used a comparative transcriptomics approach to identify genes that are uniquely overexpressed in the LFS GBM patient relative to a cancer compendium of 12,747 tumor RNA sequencing data sets, including 200 GBMs. STAT1 and STAT2 were identified as being significantly overexpressed in the LFS patient, indicating ruxolitinib, a Janus kinase 1 and 2 inhibitors, as a potential therapy. The LFS patient had the highest level of STAT1 and STAT2 expression in an institutional high-grade glioma cohort of 45 patients, further supporting the cancer compendium results. To empirically validate the comparative transcriptomics pipeline, we used a combination of adherent and organoid cell culture techniques, including ex vivo patient-derived organoids (PDOs) from four patient-derived cell lines, including the LFS patient. STAT1 and STAT2 expression levels in the four patient-derived cells correlated with levels identified in the respective parent tumors. In both adherent and organoid cultures, cells from the LFS patient were among the most sensitive to ruxolitinib compared to patient-derived cells with lower STAT1 and STAT2 expression levels. A spheroid-based drug screening assay (3D-PREDICT) was performed and used to identify further therapeutic targets. Two targeted therapies were selected for the patient of interest and resulted in radiographic disease stability. This manuscript supports the use of comparative transcriptomics to identify personalized therapeutic targets in a functional precision medicine platform for malignant brain tumors.

INNV-14. PREDICTIVE UTILITY OF TEMOZOLOMIDE RESPONSE AS DETERMINED BY 3D EX-VIVO CELL CULTURE ASSAYS IN NEWLY DIAGNOSED GLIOBLASTOMA: A SINGLE INSTITUTION CASE SERIES EVALUATING PROGRESSION FREE SURVIVAL

Glioblastoma is an aggressive tumor that is clinically and pathologically heterogeneous, and as such, remains challenging to treat. Response to initial standard treatment is widely variable and patient specific. We describe a cohort of 11 newly diagnosed patients enrolled in the 3D-PREDICT study at a single institution who underwent maximal safe surgical resection with tumor tissue collected prospectively for ex vivo cell culture assays against a panel of commonly used agents, including temozolomide. Nine of eleven patients received concurrent radiation + temozolomide followed by adjuvant temozolomide as per current standard of care. Two patients progressed immediately following concurrent chemoradiation; pseudoprogression was ruled out with short interval repeat imaging and clinical deterioration with pathology confirming recurrent tumor on re-resection in one patient. Median follow-up was 10 months (range 2-18). Outcomes were assessed retrospectively using Kaplan-Meier time-to-event curves, separating into two groups the temozolomide responders (n=4) and temozolomide nonresponders (n=7), as defined by the previously validated assay. The event was reached when radiographic tumor recurrence/progression occurred. Due to limited sample size, statistical significance was not reached, but a trend toward longer time to recurrence was noted among the temozolomide responder group. Continued experience with this tool may help clinicians predict which patients with newly diagnosed glioblastoma will respond well to initial treatment with temozolomide and those that may be more appropriate for clinical trial enrollment, independent of MGMT promoter methylation status. Longer term studies with a larger number of patients will help to determine the true significance of this drug response prediction assay.

Prospective prediction of clinical drug response in high-grade gliomas using an ex vivo 3D cell culture assay

Background

Clinical outcomes in high-grade glioma (HGG) have remained relatively unchanged over the last 3 decades with only modest increases in overall survival. Despite the validation of biomarkers to classify treatment response, most newly diagnosed (ND) patients receive the same treatment regimen. This study aimed to determine whether a prospective functional assay that provides a direct, live tumor cell-based drug response prediction specific for each patient could accurately predict clinical drug response prior to treatment.

Methods

A modified 3D cell culture assay was validated to establish baseline parameters including drug concentrations, timing, and reproducibility. Live tumor tissue from HGG patients were tested in the assay to establish response parameters. Clinical correlation was determined between prospective ex vivo response and clinical response in ND HGG patients enrolled in 3D-PREDICT (ClinicalTrials.gov Identifier: NCT03561207). Clinical case studies were examined for relapsed HGG patients enrolled on 3D-PREDICT, prospectively assayed for ex vivo drug response, and monitored for follow-up.

Results

Absent biomarker stratification, the test accurately predicted clinical response/nonresponse to temozolomide in 17/20 (85%, P = .007) ND patients within 7 days of their surgery, prior to treatment initiation. Test-predicted responders had a median overall survival post-surgery of 11.6 months compared to 5.9 months for test-predicted nonresponders (P = .0376). Case studies provided examples of the clinical utility of the assay predictions and their impact upon treatment decisions resulting in positive clinical outcomes.

Conclusion

This study both validates the developed assay analytically and clinically and provides case studies of its implementation in clinical practice.

Analytical validation of the Target Selector ctDNA platform featuring single copy detection sensitivity for clinically actionable EGFR, BRAF, and KRAS mutations

BACKGROUND

Personalized medicine requires accurate molecular profiling for targeted therapy decisions. Insufficient tissue yield or tumor heterogeneity frequently limits the correct tissue biomarker determination. As a noninvasive complement to traditional tissue biopsies, liquid biopsies detect and track cancer driver mutations from biofluids (e.g., blood, urine). Here we present the analytical validation of Target Selector™ ctDNA assays capable of single mutant DNA copy detection.

METHODS

The Target Selector ctDNA assay applies a patented Switch-Blocker technology to suppress amplification of background (wild-type) WT alleles, while allowing specific amplification of very low frequency mutant alleles. In contrast to allele specific enrichment technologies like ddPCR, one Switch-Blocker inhibits amplification of a DNA target up to 15 bp in length (e.g., one Switch-Blocker covers all KRAS exon 2, codon 12 and 13 variants). Target enrichment is achieved through a quantitative PCR reaction; subsequent DNA sequencing confirms mutation identity. Analytical validation with cancer cell line DNA was conducted by three independent operators using five instruments across five days.

RESULTS

A total of 3086 samples were tested on EGFR, BRAF and KRAS Target Selector ctDNA assays, with EGFR WT as a reference. All assays showed >99% analytical sensitivity and specificity. Single mutant copy detection is confirmed by experimental data and theoretical estimates. In the presence of 14000 WT DNA copies, limits of detection were: EGFR Del19, 0.01%; EGFR L858R, 0.02%; EGFR T790M, 0.01%; BRAF V600E, 0.01%; KRAS G12C, 0.02%. Inter- and intra-assay analyses showed r2>0.94, suggesting consistent performance among operational variables. Healthy donor samples (100 tests) showed clinical specificity at >99%. Finally, Target Selector clinical experience data of >2200 patient samples is consistent with published tissue mutation prevalence.

CONCLUSIONS

Highly sensitive Target Selector ctDNA assays with single mutant copy detection and limit of detection at 0.02% or better enable accurate molecular profiling vital for disease management.

Abstract 5597: TargetSelector™ CTC technology demonstrates clinical utility in monitoring treatment response in germ cell tumors (testicular cancer)

Background: Liquid biopsy has recently emerged as a minimally invasive and cost effective means to assess cancer biomarkers without the risk of surgical biopsy complications. Using a simple blood draw, circulating tumor cell (CTC) analysis is capable of providing information with implications for diagnostics, prognosis, and treatment decisions of various carcinomas. Serial CTC measurements can aid monitoring disease progression or response to therapy. Biocept's proprietary TargetSelector™ technology captures CTCs and provides assessment of enumeration, protein expression by immunofluorescence, and gene amplifications or translocations via FISH. There is limited data on clinical utility of CTCs in germ cell tumors (GCT). We aim to evaluate the prognostic and predictive significance of CTCs in refractory GCT in an ongoing clinical trial of brentuximab vedontin and bevacizumab in refractory CD 30+ GCT (NCT02988843). CTC enumeration will be performed at baseline, prior to cycle 3 of treatment, and at progression to monitor treatment response in blood from patients with refractory testicular cancer. Methods: Peripheral whole blood samples from refractory testicular cancer patients are collected into Biocept CEE-Sure™ blood collection tubes that are validated to preserve CTCs for up to 96 hours. Biocept's TargetSelector™ platform utilizes a proprietary antibody capture cocktail and microchannel enabling enrichment, enumeration, and CTC analyses which were performed at Biocept's CLIA-certified and CAP accredited laboratory. Results: From a patient with heavily pretreated refractory testicular cancer, serial blood collections were obtained at baseline and prior to cycle 3 (approximately six week interval). Two CTCs were detected at baseline. Prior to cycle 3, there was 1 CTC (corresponding to tumor marker response and CT scan showing stable disease). At progression after cycle 4, the CTC count increased to 4, corresponding with disease progression in tumor markers and CT scan. In this patient, the CTC capture utilizing the TargetSelector™ CTC technology was very predictive of clinical response and progression. Conclusions: Clinical application of Biocept's TargetSelector™ CTC technology enables the sensitive detection of CTCs in testicular cancer. Longitudinal CTC assessment implemented in the clinical setting can be used to assess drug responsiveness and follow disease progression in this rare disease, as well as in more common solid tumor cancers. The TargetSelector™ liquid biopsy platform provides an economical, non-invasive, and reliable means to arm physicians with valuable information for disease management and patient care. Ongoing collection of blood samples and data analysis from patients being enrolled on the study will provide us a better understanding of the potential clinical use of this novel technology in GCT patients.

Citation Format: Shilpa Gupta, Benjamin L. Maughan, Cecile Rose T. Vibat, Veena M. Singh. TargetSelector™ CTC technology demonstrates clinical utility in monitoring treatment response in germ cell tumors (testicular cancer) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 5597.

Detecting an ALK Rearrangement via Liquid Biopsy Enabled a Targeted Therapy-based Approach for Treating a Patient with Advanced Non-small Cell Lung Cancer

Initial diagnostic biopsy procedures often yield insufficient tissue for molecular testing, and invasive surgical biopsies can be associated with significant cost as well as risk to the patient. Liquid biopsy offers an alternative and economical means for molecular characterization of tumors via a simple peripheral blood draw. This case report describes the ability of liquid biopsy to detect an ALK translocation where tissue analysis by fluorescence in situ hybridization was negative for the genetic alteration. Identification of an ALK rearrangement in circulating tumor cells from a blood specimen led to sequential targeted therapies that included crizotinib followed by alectinib. The patient demonstrated outstanding clinical response during treatment with each of the prescribed ALK inhibitors. This case demonstrates the clinical utility of Biocept’s liquid biopsy to detect actionable biomarkers by surveying the systemic landscape of a patient’s disease where identification of the same genetic drivers may be missed in analyses of heterogeneous tumor tissue.

Mutation-Enrichment Next-Generation Sequencing for Quantitative Detection of KRAS Mutations in Urine Cell-Free DNA from Patients with Advanced Cancers

Purpose: Tumor-derived cell-free DNA (cfDNA) from urine of patients with cancer offers noninvasive biological material for detection of cancer-related molecular abnormalities such as mutations in Exon 2 of KRASExperimental Design: A quantitative, mutation-enrichment next-generation sequencing test for detecting KRASG12/G13 mutations in urine cfDNA was developed, and results were compared with clinical testing of archival tumor tissue and plasma cfDNA from patients with advanced cancer.Results: With 90 to 110 mL of urine, the KRASG12/G13 cfDNA test had an analytical sensitivity of 0.002% to 0.006% mutant copies in wild-type background. In 71 patients, the concordance between urine cfDNA and tumor was 73% (sensitivity, 63%; specificity, 96%) for all patients and 89% (sensitivity, 80%; specificity, 100%) for patients with urine samples of 90 to 110 mL. Patients had significantly fewer KRASG12/G13 copies in urine cfDNA during systemic therapy than at baseline or disease progression (P = 0.002). Compared with no changes or increases in urine cfDNA KRASG12/G13 copies during therapy, decreases in these measures were associated with longer median time to treatment failure (P = 0.03).Conclusions: A quantitative, mutation-enrichment next-generation sequencing test for detecting KRASG12/G13 mutations in urine cfDNA had good concordance with testing of archival tumor tissue. Changes in mutated urine cfDNA were associated with time to treatment failure. Clin Cancer Res; 23(14); 3657-66. ©2017 AACR.

Performance and Diagnostic Accuracy of a Urine-Based Human Papillomavirus Assay in a Referral Population

Background: Human papillomavirus (HPV) testing from clinician-collected cervical and self-collected cervico-vaginal samples is more sensitive for detecting CIN2+/CIN3+ than cytology-based screening, stimulating interest in HPV testing from urine. The objective was to determine the performance of the Trovagene HPV test for the detection of CIN2+ from urine and PreservCyt cervical samples.Methods: Women referred for colposcopy at St Mary's Hospital (London, United Kingdom), following abnormal cytology, were recruited to this diagnostic accuracy study by convenience sampling (September 2011 to April 2013). A total of 501 paired urine and cervical samples were collected. Primary outcomes were sensitivity for CIN2+/CIN3+ and specificity for Collapse

Key Words Collapse

MESH Headings

• Adult
• Aged
• Cervix Uteri/pathology
• Colposcopy
• DNA, Viral/isolation & purification
• Early Detection of Cancer/methods
• Feasibility Studies
• Female
• Humans
• Mass Screening/methods
• Middle Aged
• Papillomaviridae/isolation & purification
• Papillomavirus Infections/diagnosis
• Papillomavirus Infections/pathology
• Papillomavirus Infections/urine
• Papillomavirus Infections/virology
• Prospective Studies
• RNA, Viral/isolation & purification
• Referral and Consultation
• Sensitivity and Specificity
• Specimen Handling
• United Kingdom
• Urinalysis/methods
• Uterine Cervical Neoplasms/diagnosis
• Uterine Cervical Neoplasms/pathology
• Uterine Cervical Neoplasms/urine
• Uterine Cervical Neoplasms/virology
• Vaginal Smears
• Uterine Cervical Dysplasia/diagnosis
• Uterine Cervical Dysplasia/pathology
• Uterine Cervical Dysplasia/urine
• Uterine Cervical Dysplasia/virology

Collapse

Grants

• 16891 Cancer Research UK
• C569/A16891 Cancer Research UK

Collapse

Affiliation(s)

Jack Cuzick Centre for Cancer Prevention, Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, United Kingdom.
Louise Cadman Centre for Cancer Prevention, Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, United Kingdom
Amar S Ahmad Centre for Cancer Prevention, Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, United Kingdom
Linda Ho Centre for Cancer Prevention, Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, United Kingdom
George Terry Centre for Cancer Prevention, Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, United Kingdom
Michelle Kleeman Centre for Cancer Prevention, Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, United Kingdom
Deirdre Lyons Imperial College Healthcare NHS Trust, St Mary's Hospital, London, United Kingdom
Janet Austin Centre for Cancer Prevention, Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, United Kingdom
Mark H Stoler Department of Pathology, University of Virginia Health System, Charlottesville, Virginia
Cecile Rose T Vibat Trovagene Inc., San Diego, California
Janel Dockter Trovagene Inc., San Diego, California
David Robbins Trovagene Inc., San Diego, California Genomind Inc. King of Prussia, Pennsylvania
Paul R Billings Trovagene Inc., San Diego, California
Mark G Erlander Trovagene Inc., San Diego, California

Collapse

Performance of Biocept's sample collection for tumor cell analysis

e23036

Background: Liquid biopsy is a minimally invasive and cost effective way to assess cancer biomarkers without the risk of surgical biopsy complications. Circulating tumor cell (CTC) analysis from body fluids can provide critical information towards early detection, prognosis and treatment decisions. Accurate CTC evaluations require optimal cell preservation. Cell lysis, DNA degradation, or membrane alterations compromise CTC analyses and accurate diagnoses. This work compares Biocept’s proprietary CEE-Sure BCT and Saccomanno's Cytology Fixative largely used for sputum collection. Methods: One million BT474 (HER2 amplified) or H3112 (ALK re-arranged) cells were spiked into 500 µl medium; 500 µl of CEE-Sure or Saccomanno fixative was added. Tubes were stored at 4°C for 1 day, 1 week, or 1 month. Cells were centrifuged, resuspended, and counted (Celigo). Around 150 cells in 15 µl of RPMI medium were flowed into Biocept's microfludic system for cell capture; recovery (%) was calculated. Captured cells were subjected to fluorescent in situ hybridization (FISH) analyses for qualitative signal evaluation. Results: As similar results were observed for both cell lines and all time points, combined data will be shown. Median cell recovery after CEE-Sure incubation was 14.1% (range 1.7–44%, n = 12) vs 5.4% (range 0.07–26.9%, n = 12) in Saccomanno's fixative. Median cell capture of ~150 cells fed into Biocept’s microchannel was 96% (range 72-98%) for CEE-Sure vs 82% (range 21-96%) for Saccomanno. Paired t-tests showed significant differences for both recovery and capture. FISH signals from CEE-Sure samples were qualitatively rated Fair to Good, while Saccomanno samples had Poor to Fair, grainy, non-specific signals. Conclusions: This preliminary work shows consistently higher cell recovery, better cell membrane maintenance, and higher quality FISH signals for samples stored in Biocept's CEE-Sure vs Saccomano’s fixative. With liquid biopsy testing gaining rapid traction, maximal cell stability during the transport and storage are crucial. Additional fixative comparison is ongoing in various patient specimen types. These results support expansion of molecular analyses in sputum samples enriched for lung epithelial cells.

ctDNA detection of EGFR mutations in NSCLC patients using TargetSelector

e23037

Background: EGFR mutations are the most frequent targetable genomic alterations in non-small cell lung cancer (NSCLC) patients (pts). While tissue biopsy remains the standard for assessing of EGFR mutation status, it is invasive and not always feasible. Liquid biopsy is a minimally invasive alternative. Biocept’s proprietary TargetSelector system evaluates circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA) in blood. We aimed to clinically validate the accuracy of EGFR-specific TargetSelector in NSCLC pts. Methods: At three time points (T0: baseline before TKI, T1: during EGFR-TKI therapy, T2: after progression), blood samples were collected in Biocept OncoCEE BCT validated to preserve DNA up to 8 days. These samples were interrogated for three EGFR mutations: exon 19 deletions (Del 19), L858R, and T790M. The objectives are to assess detection sensitivity of liquid biopsy using EGFR mutation status vs the tissue as gold standard and to evaluate whether the detection sensitivity changes with EGFR-TKI therapy. Results: A total of 53 study pts were enrolled (male, 21; female, 32). The mean age was 70.6 (range: 46 – 90). Most pts had stage IV disease (43, 81.1%) and lung adenocarcinoma (48, 90.6%). 26 (49.1%) pts had EGFR mutations in tumor tissue: Del 19, 13; L858R, 8; T790M, 6; other, 8. Detection sensitivity for sensitizing EGFR mutations (Del 19 and L858R) at T0, T1, and T2 was 60.0% (6/10), 33.3% (5/15), and 33.3% (1/3), respectively. There was no statistical difference in CTC counts between activating EGFR mutation-positive and -negative pts (mean CTC count: 10.5 vs 20.1; p = 0.11 by two-sided t-test). Detection sensitivity for T790M was 33.3% (2/6) and 5 of 6 pts were receiving T790M directed therapy (3, rociletinib; 2, osimertinib) at the time of blood draw. Two pts – one patient before initiation of EGFR-TKI and the other during treatment with erlotinib – were found to have T790M mutations only in blood and not in tissue. Conclusions: Activating EGFR mutation detection may decrease during the course of TKI therapy, possibly due to treatment response. Further research with an expanded sample size and serial collections are needed to evaluate this finding, and to investigate possible implications of the presence of T790M only in blood.

Sensitivity of TargetSelector in clinical experience in ctDNA profiling of NSCLC 2000 cases

e23033

Background: Targeted cancer therapy relies on identifying specific DNA mutations from a patient’s tumor. Tyrosine kinase inhibitors (TKIs) tend to be effective for non-small cell lung cancer (NSCLC) with epidermal growth factor receptor (EGFR) activating mutations, of which exon 19 deletions (Del19) and L858R are most common. Acquired resistance to TKI therapy is associated with a T790M mutation. Standard biomarker analyses may not reflect tumor heterogeneity; they entail tissue biopsies often with surgical complications. To address these limitations, Biocept developed a minimally invasive method to characterize cancer biomarkers in blood. Biocept's proprietary TargetSelector assays selectively amplify relevant mutations from circulating tumor DNA (ctDNA). Clinical validations demonstrated high concordances between molecular tests in blood vs tissue. As further validation, EGFR mutation detection frequencies were compared to US averages (mycancergenome.org). Here we analyze 2000 blood samples received at Biocept from 1Mar 2016 to 4Jan 2017 from late stage NSCLC patients. Methods: Blood was collected in Biocept OncoCEE BCT validated to preserve DNA ≤ 8 days. TargetSelector was used to detect ctDNA L858R, Del19 and T790M.EGFR allele copy numbers for wild type and each mutant were calculated. The prevalence of each mutation was compared to US averages. Results: Del19, L858R, and T790M mutations were detected in 12.9%, 8.5%, and 9.9% of the analyzed blood samples, respectively. This is concordant with US averages, which are 10% for each mutation. Median copy numbers/ml of blood were 30 for Del19 (range: 1 – 91974), 15 for L858R (range: 1 – 91200), and 10 for T790M (range: 1 – 137360). The median wild type EGFR copy number detected/ml blood was 2304 (range: 8 – 2498725). In ~80% of T790M cases, ≥ 1 concomitant activating mutation was detected. Conclusions: Biocept's TargetSelector detects EGFR mutations (Del19, L585R, and T790M) at a very high level of sensitivity down to 1 mutant copy/ml in advanced NSCLC patients at frequencies consistent with cited US rates. Moreover, the underlying activating mutation was detected in ~80% of T790M cases.

Abstract 3146: Circulating tumor DNA assay performance for detection and monitoring of KRAS mutations in urine from patients with advanced cancers

Introduction: Non-invasive urinary ctDNA-based liquid biopsy approach can be used to detect and track cancer driver mutations for rapid diagnosis and disease monitoring. Using highly sensitivity ctDNA mutation detection platform, we examined the detection of KRAS G12/13 mutations in urine obtained from advanced cancer patients, assessed urine sample requirements, and compared the results with matched tumor tissue in patients with advanced cancers.

Methods: 41 patients with advanced solid cancer with KRAS mutations on archival tumor from CLIA laboratory testing were prospectively enrolled with informed consent (colorectal cancer, n = 29; non-small cell lung cancer, n = 6; pancreatic cancer, n = 2; ovarian cancer, n = 2; other, n = 2). Urine was collected before and during experimental therapies. Urinary DNA was isolated using a method that enriches for highly fragmented, systemically derived cell-free DNA. KRAS G12/13 analysis was performed using mutation enrichment PCR coupled with next generation sequencing (MiSeq). Analytical sensitivity of the KRAS G12/13 assay is 0.006% mutant alleles in the background of 60 ng wild-type (wt) DNA and 0.002% mutant alleles in 360 ng wt DNA. Clinical data was collected retrospectively from the electronic medical record.

Results: For 41 patients enrolled on a study, urine volumes in pretreatment samples ranged from 13 to 120 mL (median, 55 mL). Urinary DNA yields were 151 to 23059 ng (median, 1039 ng). Using tissue as the reference, the positive percent agreement (PPA) between urine and tumor KRAS G12/13 test results was 54% (22/41) for urine samples with all volumes (13-120 mL) and any DNA input amount (2-360 ng) and 92% (12/13) for urine samples with volumes ≥50 mL and DNA input amount ≥60 ng. For metastatic CRC patient cohort, the PPA between urine and tumor KRAS G12/13 test result was 60% (18/30) for urine samples with all volumes and any DNA input amount (20-120 mL, 2-360 ng) and 100% (10/10) for urine samples with volumes ≥50 mL and DNA input amount ≥60 ng. Feasibility of longitudinal monitoring KRAS G12/13 mutational burden in urine of patients treated with experimental therapies was demonstrated.

Conclusion: KRAS G12/13 mutational status can be assess in urinary DNA with highest PPA amongst patients with urine volume ≥50 mL and DNA input amount ≥60 ng (92%). KRAS mutation detection from urine should be considered as a viable approach, particularly when tumor tissue is not available.

Citation Format: Takeo Fujii, Cecile Rose T. Vibat, Daniel D. Karp, Sarina A. Piha-Paul, Vivek Subbiah, Apostolia M. Tsimberidou, Siquing Fu, David S. Hong, Helen J. Huang, Kiran Madwani, Debra L. Andrews, Saege Hancock, Aung Naing, Rajyalakshmi Luthra, Bryan K. Kee, Scott Kopetz, Mark G. Erlander, Vlada Melnikova, Funda Meric-Bernstam, Filip Janku. Circulating tumor DNA assay performance for detection and monitoring of KRAS mutations in urine from patients with advanced cancers. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 3146.

Abstract A31: Assessment of EGFR mutations in matched urine, plasma and tumor tissue in NSCLC patients treated with rociletinib (CO-1686)

Background: The acquisition of suitable tumor tissue is a challenge for a significant fraction of late-stage NSCLC patients who require EGFR testing to inform choice of therapy. An alternative for these patients could be the assessment of EGFR mutations in circulating tumor DNA (ctDNA). In this study, we examined the detection of EGFR T790M mutation in ctDNA from urine, assessed urine sample requirements, and compared the results with contemporaneously matched tumor tissue and plasma in TIGER-X (NCT01526928), a Phase 1/2 clinical study of rociletinib in previously treated mutant EGFR patients with advanced NSCLC. Rociletinib is an oral, potent, small-molecule irreversible tyrosine kinase inhibitor that selectively targets mutant forms of EGFR, including T790M, L858R and Del(19), while sparing wild-type EGFR.

Methods: 63 Stage IIIB/IV NSCLC patients enrolled in either Phase 1 or 2 components of TIGER-X and representing all therapeutic dose groups consented to optional urine collection. Maximum sample volumes were 100 mL for urine and 2 mL for plasma. To maximize assay sensitivity in urine, samples containing the recommended sample volume of ≥90 mL (≥ 90% of maximum in this study) were evaluated; all samples received were processed to assess this recommendation. Urinary and plasma ctDNA were tested for mutations by the same EGFR assays using a sensitive and quantitative short footprint assay method that employs a mutation enrichment step followed by next generation sequencing.

Results: Urine volumes ranged from 8-100 mL with a median DNA yield of 313 ng (N = 63). The median DNA yield was 299 ng for urine specimens with volume &lt;90 mL (N = 44) and 392 ng for specimens with the previously recommended volume of ≥90 mL (N = 19). 45/63 patients (71%) were positive for T790M by urine in comparison to 47/63 (75%) by central lab testing of tissue. Using tissue as the reference, the positive percent agreement (PPA) between urine and tumor T790M test results was 93% (13/14) using urine samples with volumes ≥90 mL and 73% (34/47) when the urine volume was &lt;90 mL. The PPA between plasma and tissue T790M test results was 80% (42/51). Overall, there were 13 cases that were tumor T790M+/urine T790M-, and 7 cases that were urine T790M+/tumor T790M-. Six urine T790M+/tumor T790M- cases were also T790M+ in plasma. Two additional urine T790M+ cases could not be analyzed by the tumor test due to insufficient material. Serially matched urine and plasma assessments were performed on a subset of patients (N = 15). In most cases, the urine and plasma serial profiles mirrored one another. A rapid drop in urine and plasma mutant EGFR levels was seen in almost all patients who experienced clinical benefit from rociletinib.

Conclusions: The analysis of ctDNA from urine identified a similar proportion of T790M+ patients as tissue-based testing with highest PPA amongst patients with urine volumes ≥90 mL. Discordant samples between urine and tissue that were not identified by the tumor test may be explained by tumor heterogeneity and/or inadequate biopsy. EGFR mutation detection from urine increases with urine volume and DNA yields and should be considered as a viable approach, particularly when tumor tissue is not available. Lastly, monitoring urine ctDNA T790M mutations longitudinally with baseline and post-therapy sampling could be clinically useful to determine benefit from therapy.

Citation Format: Shirish Gadgeel, Chris Karlovich, Vlada Melnikova, Lecia V. Sequist, D. Ross Camidge, Heather Wakelee, Maurice Perol, Geoffrey R. Oxnard, Karena Kosco, Cecile Rose T. Vibat, Elaina Mann, Shannon Matheny, Lindsey Rolfe, Mitch Raponi, Mark G. Erlander, Karen Reckamp. Assessment of EGFR mutations in matched urine, plasma and tumor tissue in NSCLC patients treated with rociletinib (CO-1686). [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr A31.

Abstract B4: Use of urinary circulating tumor DNA (ctDNA) KRAS for monitoring treatment response in patients with metastatic colorectal cancer (mCRC)

Background: Radiographic imaging in patients with mCRC is the current standard of care for monitoring responses. At least 40% of patients with mCRC have tumor associated KRAS mutations, and these may be detected in tumor DNA in plasma and urine. The aim of this study was to correlate the dynamics of KRAS mutational load in urinary and plasma ctDNA with clinical responses in patients with mCRC.

Methods: In a blinded biomarker study of 30 metastatic patients CRC receiving chemotherapy or surgical treatment for resectable liver metastases, interim analysis was conducted in 13 patients (7 harboring KRAS G12/13 mutation in the tumor tissue, 6 wild-type for KRAS by tissue). For KRAS-positive patients, urine and plasma specimens were obtained pre-treatment, 2 weeks after initiation of chemotherapy, and subsequently at monthly intervals. Urine ctDNA was extracted using a method that preferentially isolates short fragmented ctDNA. Quantitative analysis of KRAS G12/13 mutation load in both urine and plasma was performed with the same validated KRAS G12/13 assay, which utilizes a wild-type blocker PCR enrichment followed by a next generation sequencing-based detection (MiSeq) with standardized reporting of mutant copies per 10^5 genome equivalents.

Results: In an interim analysis, patients enrolled had pre-treatment urine and plasma specimens. In addition, a subset of patients (6) had longitudinal collections for up to 17 months. Eleven of 13 patients had liver-dominant metastatic disease. All KRAS tissue positive patients were receiving FOLFOX; all KRAS tissue-negative patients were receiving FOLFIRI plus cetuximab. In all KRAS G12/13 tissue positive patients with liver metastatic disease, a urinary ctDNA KRAS mutation concordant with tissue was identified; in 2 patients with non-liver metastatic disease, KRAS mutation was discordant or unidentifiable suggesting different biology in these patients. Plasma ctDNA KRAS mutation concordant with tissue was identified in 6 of 7 patient. Overall, the dynamics of ctDNA KRAS mutation burden in urine showed stronger concordance with clinical course and lack of temporal fluctuations as compared to the dynamics of ctDNA KRAS in plasma. A significant decrease in urinary ctDNA KRAS signal was observed as early as 2 weeks on chemotherapy, and this molecular response correlated with subsequently documented radiographic response. In 1 patient with documented progressive disease, an increase in urinary ctDNA KRAS G12/13 signal was observed 2 months prior to radiographic progression. Finally, 5 of 6 patients who tested negative for KRAS G12/13 mutations by tumor biopsy also tested negative in urine and plasma; 1 additional tissue KRAS wild-type patient tested positive for KRAS G12D mutation by urine and plasma ctDNA.

Conclusions: In a proof of concept study, urine ctDNA was shown to carry a concordant KRAS mutation with tissue of mCRC patients. ctDNA analysis identified additional KRAS G12/13 mutation positive patient in whom KRAS mutations were undetected by the tumor test suggesting tumor heterogeneity and/or inadequate biopsy. The dynamics of ctDNA KRAS mutational load in urine correlated with radiographic responses, especially in liver-dominant metastatic patients, suggesting that urinary ctDNA may be a valuable method for monitoring treatment responses in patients with mCRC.

Citation Format: Afsaneh Barzi, Vlada Melnikova, Cecile Rose T. Vibat, Saege Hancock, Yan Ning, Dana Agafitei, Mark G. Erlander, Heinz-Josef Lenz. Use of urinary circulating tumor DNA (ctDNA) KRAS for monitoring treatment response in patients with metastatic colorectal cancer (mCRC). [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2015 Nov 5-9; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2015;14(12 Suppl 2):Abstract nr B4.

Comparison of urine specimen collection times and testing fractions for the detection of high-risk human papillomavirus and high-grade cervical precancer

BACKGROUND

Urine testing for high-risk human papillomavirus (HR-HPV) detection could provide a non-invasive, simple method for cervical cancer screening.

OBJECTIVES

We examined whether HR-HPV detection is affected by urine collection time, portion of urine stream, or urine fraction tested, and assessed the performance of HR-HPV testing in urine for detection of cervical intraepithelial neoplasia grade II or worse (CIN2+).

STUDY DESIGN

A total of 37 female colposcopy clinic attendees, ≥ 30 years, provided three urine samples: "first void" urine collected at home, and "initial stream" and "mid-stream" urine samples collected at the clinic later in the day. Self- and physician-collected brush specimens were obtained at the same clinic visit. Colposcopy was performed and directed biopsies obtained if clinically indicated. For each urine sample, HR-HPV DNA testing was conducted for unfractionated, pellet, and supernatant fractions using the Trovagene test. HR-HPV mRNA testing was performed on brush specimens using the Aptima HPV assay.

RESULTS

HR-HPV prevalence was similar in unfractionated and pellet fractions of all urine samples. For supernatant urine fractions, HR-HPV prevalence appeared lower in mid-stream urine (56.8%[40.8-72.7%]) than in initial stream urine (75.7%[61.9-89.5%]). Sensitivity of CIN2+ detection was identical for initial stream urine and physician-collected cervical specimen (89.9%[95%CI=62.7-99.6%]), and similar to self-collected vaginal specimen (79.1%[48.1-96.6%]).

CONCLUSION

This is among the first studies to compare methodologies for collection and processing of urine for HR-HPV detection. HR-HPV prevalence was similar in first void and initial stream urine, and was highly sensitive for CIN2+ detection. Additional research in a larger and general screening population is needed.

Abstract 5237: Monitoring minimal residual disease by urinary or plasma circulating tumor DNA of KRAS mutation burden in colorectal cancer patients with resectable liver metastases

Background: Colorectal cancer (CRC) is the third most commonly diagnosed cancer and the third leading cause of cancer deaths. Over half of patients with CRC will develop liver metastases. Surgical resection, in combination with systemic therapies, greatly improves long-term outcomes, and around 40% of patients with resected liver limited disease are alive 5 years after diagnosis. While tumor staging and radicality of surgery are commonly used for prognostic assessment, better non-invasive markers are needed for monitoring chemo-responsiveness, following minimal residual disease (MRD), and guiding complex treatment decisions in these patients. This study evaluated the utility of quantitating KRAS mutation burden in urinary and plasma ctDNA as a means of monitoring MRD in surgical CRC patients with liver limited metastases.

Methods: We developed a novel, NGS-based method for enrichment and quantitative detection of KRAS mutations in urinary and plasma ctDNA. A blinded retrospective study was conducted on 20 Stage I-IV CRC patients, 15 of whom had undergone curative or palliative intent surgical resection of primary tumor or liver metastases in combination with various systemic therapies.

Results: Archived, matched urine and plasma samples were collected from 20 patients with KRAS positive primary tumor. For operable patients, specimens were collected prior to surgery, during and immediately after surgery, plus additional time points post-surgery. A total of 193 plasma and urine samples archived for 3-5 years were tested. All 101 of 101 plasma samples (100%) and 79 of 92 urine samples (86%) had sufficient DNA and were deemed evaluable. In a blinded analysis, a correct KRAS mutation that correlated with KRAS mutation in tissue was identified in 95% of evaluable baseline plasmas (19 of 20) and 92% of evaluable baseline urines (11 of 12). In one patient, a KRAS mutation distinct from that identified in the tumor biopsy was detected consistently across all serial ctDNA samples. Overall, we observed a clear correlation and highly comparable fold change between plasma and urinary ctDNA KRAS levels on treatment. Significantly, in all patients with curative intent surgery, ctDNA KRAS levels were undetectable in urine or plasma after surgery. In contrast, in 8 of 10 patients with incomplete, palliative surgery, the ctDNA KRAS signal remained detectable or increased after surgery. Further correlation between ctDNA KRAS and clinical outcomes will be discussed.

Conclusion: We demonstrate for the first time that quantitative changes of mutational KRAS burden in plasma and urinary ctDNA are highly correlated. We further demonstrate clinical applicability of urinary ctDNA KRAS analysis for monitoring quantitatively, with single molecule sensitivity, the MRD post-surgery for CRC patients with liver resectable metastases. Supported by Grant no. 13660.

Citation Format: Vlada Melnikova, Jason C. Poole, Cecile Rose T. Vibat, Lucie Benesova, Barbora Belsanova, Saege Hancock, Latifa Hassaine, Errin Samuelsz, Timothy T. Lu, Mark G. Erlander, Marek Minarik. Monitoring minimal residual disease by urinary or plasma circulating tumor DNA of KRAS mutation burden in colorectal cancer patients with resectable liver metastases. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5237. doi:10.1158/1538-7445.AM2015-5237

Abstract 5240: Comparative levels of KRAS mutations circulating tumor DNA for association with overall survival in patients with non-resectable pancreatic cancer

Background: The median overall survival (OS) time of patients with non-resectable pancreatic cancer varies widely. Diagnostic tools are presently lacking to predict patient outcome at diagnosis. The vast majority of pancreatic tumors harbor KRAS mutations. In this study, we evaluated whether quantitative baseline and longitudinal monitoring of KRAS mutations in plasma circulating tumor DNA (ctDNA) may be used to stratify patients for predicting outcome.

Methods: The Danish BIOPAC study prospectively collected plasma from patients with non-resectable pancreatic cancer undergoing treatment with gemcitabine or FOLFIRINOX. Archival (3-5 years) plasma specimens were collected from 113 patients pre-treatment (baseline),on chemotherapy, as well as at multiple additional time intervals for up to 977 days from baseline. Interim analysis of ctDNA KRAS was conducted (after 105 deaths). Levels of ctDNA KRAS mutations were assessed in 35 patients with long OS (median 473 days; range 360-1134), 33 patients with medium OS (median 227 days; range 155-349) and 37 patients with short OS (median 94 days; range 21-146). PCR enrichment of KRAS G12A/C/D/R/S/V, and G13D mutations was performed, followed by massively parallel deep sequencing and quantification with standardization of reporting number of copies detected per 105 genome equivalents (GE).

Results: In a prospective-retrospective biomarker study of 113 patients, interim analysis of ctDNA KRAS was conducted (after 105 deaths). 92 of 105 patients had evaluable baseline plasma samples. Number of mutant KRAS copies was higher in patients with short OS (median 661; range 0-190,490 copies/105 GE) versus with median OS (median 103; range 0 to 275,918 copies/105 GE) versus with long OS (median, 15; range, 0-1,369 copies/105 GE). Longitudinally, KRAS mutation levels remained mostly low with long OS (last time point median 9; range 0-70,451 copies/105 GE) vs. medium OS (median 155; range 0-314,103 copies/105 GE) or short OS where levels increased or remained high (median 803; range 0-138,508 copies/105 GE). As this dramatic difference in systemic KRAS levels may reflect distinct tumor phenotypes, the underlying tumor biology was further investigated by interrogating additional cancer mutational hotspots (using massively parallel deep sequencing) in plasma ctDNA of patients stratified by systemic KRAS and the OS.

Conclusion: Shorter OS in patients with non-resectable pancreatic cancer tended to associate with high levels of ctDNA KRAS mutations at diagnosis and with post-treatment elevation of KRAS mutations. ctDNA KRAS mutation levels in patients with non-resectable pancreatic cancer observed at diagnosis or on treatment may predict patient outcome and could reflect distinct underlying tumor biology.

Citation Format: Julia S. Johansen, Cecile Rose T. Vibat, Saege Hancock, Latifa Hassaine, Errin Samuelsz, Inna Chen, Eric A. Collisson, Dan Calatayud, Benny V. Jensen, Jane Preuss Hasselby, Timothy T. Lu, Jason C. Poole, Vlada Melnikova, Mark G. Erlander. Comparative levels of KRAS mutations circulating tumor DNA for association with overall survival in patients with non-resectable pancreatic cancer. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5240. doi:10.1158/1538-7445.AM2015-5240

Abstract 5238: Methodology for single copy detection and quantitative monitoring of clinically actionable circulating tumor DNA mutations in urine from cancer patients

Background: Non-invasive detection and monitoring of circulating tumor DNA (ctDNA) mutations for personalized treatment of cancer patients can be realized by combining the practical advantages of urine as a ctDNA sample source with high throughput of next-generation sequencing (NGS).

Methods: Our platform couples an extraction process capable of isolating ctDNA from the entire void volume of a urine sample (∼100ml) with an ultra-sensitive NGS-integrated mutation enrichment method with single copy detection sensitivity. Assays have been developed and validated to interrogate clinically actionable mutations/deletions in the KRAS, BRAF and EGFR (Exons 19, 20, 21) oncogenes in both urine and plasma samples. For mutation detection with high sensitivity, a novel allele-specific competitive cycling (ASCC) method was used prior to NGS to amplify ultra-short target DNA (31-45 bp) using kinetically-favorable binding conditions for a wild type (WT) blocking oligonucleotide. Enriched amplicons were sequenced and a proprietary algorithm was used to quantify the mutant ctDNA input level in analytical and clinical samples.

Results: An extraction method optimized for enrichment of fragmented urine ctDNA enabled isolation of large amounts of DNA (mean ∼2 μg) from advanced stage patients with different cancer types. Analytical performance characterization demonstrated sensitivity of 0.0047 - 0.01% mutant copies in WT/mutant DNA blends. To show that our assays have a true single copy detection threshold, we used DNA blends with defined mutant spike-in levels of 2 - 60 copies distributed over 20 wells (to obtain 0 - 3 mutant copies/well). The observed distribution of positive and negative hits matched the theoretical hit rate of an ideal Poisson distribution for these replicates, confirming single copy sensitivity of our assays. Using the KRAS ASCC assay, 1 to 17,555 mutant fragments were detected per 1 mL of urine collected from KRAS tissue biopsy-positive advanced cancer patients. Analysis of serial patient-matched urine and plasma longitudinal samples from KRAS, BRAF, or EGFR tissue positive patients demonstrated a high level of concordance between urine and plasma samples and feasibility of monitoring mutation load in a variety of clinical settings, including monitoring post-surgery and responsiveness to targeted therapy or chemotherapy.

Conclusion: The analytical characterization and clinical feasibility studies demonstrate that this methodology can successfully detect and quantitate mutational load in urinary ctDNA, thus enabling for the dynamic monitoring of therapy response, drug resistance, and minimal residual disease in cancer patients from a truly non-invasive sample.

Citation Format: Karena Kosco, Jason C. Poole, Saege Hancock, Errin Samuelsz, Timothy T. Lu, Erin Clark, Latifa Hassaine, Shiloh Guerrero, Cecile Rose T. Vibat, Vlada Melnikova, Mark G. Erlander. Methodology for single copy detection and quantitative monitoring of clinically actionable circulating tumor DNA mutations in urine from cancer patients. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5238. doi:10.1158/1538-7445.AM2015-5238

BRAF V600E mutations in urine and plasma cell-free DNA from patients with Erdheim-Chester disease

Erdheim-Chester disease (ECD) is a rare histiocytosis with a high prevalence of BRAF V600E mutation (>50% of patients). Patients with BRAF-mutant ECD can respond to BRAF inhibitors. Unfortunately, the lack of adequate archival tissue often precludes BRAF testing. We hypothesized that cell-free DNA (cfDNA) from plasma or urine can offer an alternative source of biologic material for testing. We tested for BRAF V600E mutation in cfDNA from the plasma and urine of 6 ECD patients. In patients with available archival tissue, the result of BRAF mutation analysis was concordant with plasma and urine cfDNA results in all 3 patients (100% agreement, kappa 1.00). In all 6 patients, BRAF mutation analysis of plasma and urine cfDNA was concordant in 5 of 6 patients (83% agreement, kappa 0.67). Testing for BRAF V600E mutation in plasma and urine cfDNA should be further investigated as an alternative to archival tissue mutation analysis.

Highly sensitive quantitative detection of circulating tumor DNA in urine and plasma from advanced colorectal cancer patients in aid of early diagnosis of clinically relevant KRAS mutations

654

Background: Acquisition of point mutations in KRAS gene is causally associated with the onset of development of a resistance to anti-EGFR therapy in colorectal cancer. Newly acquired KRAS mutations can be detected in blood plasma months before radiographic detection. The objective of this study was to demonstrate feasibility of an ultrasensitive non-invasive method for detection of KRAS mutations in urine and plasma of patients with advanced colorectal cancer. Methods: Archived, matched urine and plasma samples (stored between 3-5 years prior to ctDNA extraction) from 20 treatment naïve, advanced stage cancer patients with known tumor tissue KRAS mutations determined by an accredited clinical laboratory, were used in a retrospective setting for a blinded concordance study. KRAS status in urine and plasma was compared to that in tumor tissue in order to assess clinical sensitivity of the ctDNA assay. An ultrashort-amplicon (31bp) assay for KRAS mutation enrichment and detection in highly fragmented urinary and plasma ctDNA was developed. The assay detected 1 copy of KRASG12A/C/D/R/S/V or G13D mutant allele in a background of wild-type DNA with a verified analytical sensitivity of 0.007% (7 copies per ~100,000 genome equivalents). Results: In a pilot study of 20 advanced stage colorectal cancer patients, 15 of 16 evaluable archived urine samples (94%) had KRAS mutation that was concordant with tissue biopsy. Of 20 archived plasma samples evaluated, 19 (95%) displayed the KRAS mutation concordant with tumor tissue. Of 16 paired urine and plasma samples, 15 (94%) had concordant KRAS mutation calls. Conclusions: This study demonstrates high clinical sensitivity (≥94%) of concordant KRAS mutation detection between urine, plasma and tissue specimens from advanced colorectal cancer patients. Early detection and monitoring of acquired KRAS mutations in circulating tumor DNA, and in particular urinary ctDNA, opens the possibility of a new paradigm for a truly non-invasive method of individualized care for colorectal cancer patients.

Comparative circulating tumor DNA levels for KRAS mutations in patients with nonresectable pancreatic cancer

288

Background: Non-resectable pancreatic cancer patients have a wide range of median time for overall survival (OS). Currently there is a lack of diagnostic tools to predict patient outcome at diagnosis. KRAS mutations are present in the vast majority of pancreatic tumors. The study objective was to determine if quantitative baseline and longitudinal monitoring of KRAS mutations from plasma circulating tumor DNA (ctDNA) could be used to stratify patients for predicting outcome. Methods: Plasma was prospectively collected from the Danish BIOPAC study for non-resectable pancreatic cancer patients undergoing treatment with gemcitabine or FOLFIRINOX. Feasibility of monitoring ctDNA KRAS mutations was assessed in 10 patients with long OS (median 493 days; range 360-1031) and 10 patients with short OS (median 66 days; range 21-136). KRAS G12A/C/D/R/S/V, and G13D mutations were PCR enriched, sequenced by massively parallel deep sequencing, quantitated and standardized by reporting number of copies detected per 105 genome equivalents (GE). Results: In a pilot study of 20 patients, all 18 patients with evaluable DNA had detectable KRAS mutations. Of 18 patients, 12 had baseline plus longitudinal time points (7 short, 5 long OS). Mutant KRAS copies were higher for short OS (median=994; range 0-34305 copies/105 GE) vs. with long OS (median 196; range, 34-278 copies/105 GE). Longitudinally, KRAS mutation levels remained mostly low with long OS (last time point median 204; range 8-873 copies/105 GE) vs. short OS where levels increased or remained high (median 2363; range 71-47919 copies/105 GE). Identical KRAS mutations were consistently detected for a given patient with short OS. However, long OS patients had variable KRAS mutations in longitudinal analysis. Conclusions: High levels of ctDNA KRAS mutations at diagnosis and post-treatment elevation of KRAS mutations were more associated with short OS. Different levels of KRAS mutation at diagnosis may predict patient outcome and could reflect distinct underlying tumor biology. Expansion of this prospective-retrospective biomarker cohort will be reported.

In vitro nanobody discovery for integral membrane protein targets

Nanobodies (Nbs) or single-domain antibodies are among the smallest and most stable binder scaffolds known. In vitro display is a powerful antibody discovery technique used worldwide. We describe the first adaptation of in vitro mRNA/cDNA display for the rapid, automatable discovery of Nbs against desired targets, and use it to discover the first ever reported nanobody against the human full-length glucose transporter, GLUT-1. We envision our streamlined method as a bench-top platform technology, in combination with various molecular evolution techniques, for expedited Nb discovery.

Prospective blinded study of BRAFV600E mutation detection in cell-free DNA of patients with systemic histiocytic disorders

Patients with Langerhans cell histiocytosis (LCH) and Erdheim-Chester disease (ECD) have a high frequency of BRAF(V600E) mutations and respond to RAF inhibitors. However, detection of mutations in tissue biopsies is particularly challenging in histiocytoses due to low tumor content and stromal contamination. We applied a droplet-digital PCR assay for quantitative detection of the BRAF(V600E) mutation in plasma and urine cell-free (cf) DNA and performed a prospective, blinded study in 30 patients with ECD/LCH. There was 100% concordance between tissue and urinary cfDNA genotype in treatment-naïve samples. cfDNA analysis facilitated identification of previously undescribed KRAS(G12S)-mutant ECD and dynamically tracked disease burden in patients treated with a variety of therapies. These results indicate that cfDNA BRAF(V600E) mutational analysis in plasma and urine provides a convenient and reliable method of detecting mutational status and can serve as a noninvasive biomarker to monitor response to therapy in LCH and ECD.

SIGNIFICANCE

Patients with BRAF(V600E)-mutant histiocytic disorders have remarkable responses to RAF inhibition, but mutation detection in tissue in these disorders is challenging. Here, we identify that analysis of plasma and urinary cfDNA provides a reliable method to detect the BRAF(V600E) mutation and monitor response to therapy in these disorders.

Abstract LB-170: Droplet digital PCR detection and longitudinal monitoring of BRAF mutations in cell-free urinary DNA of patients with metastatic cancers or Erdheim-Chester disease

Background: Detection and monitoring of oncogenic mutations in cell-free urinary DNA opens the possibility of a new paradigm for a truly non-invasive method of individualized care for metastatic cancer patients, which would enable the quantitation of mutational tumor load and respective concordance to therapeutic responsiveness followed by detection of emerging genomic alterations underlying acquired resistance.

Methods: Cell-free DNA was isolated from single and/or multiple sequential urine samples from patients with advanced cancers or Erdheim-Chester disease and BRAF V600E mutation in the tumor tissue from a CLIA-certified laboratory, who progressed on systemic therapy. Assays for quantitative assessment of BRAF V600E mutation in cell-free urinary DNA were developed using digital droplet PCR methodology (RainDance, MA) with enrichment of mutation-containing DNA fragments by pre-amplification of the BRAF gene. Detection limits were established as wild-type (&lt;0.05% of mutant copies), low-mutant (0.05%-0.107%), mutant (&gt;0.107%).

Results: Cell-free DNA was extracted from urine of 33 patients with diverse advanced cancers (melanoma, n=11; colorectal cancer, n=8; non-small cell lung cancer, n=4; papillary thyroid carcinoma, n=4; other, n=4) or Erdheim-Chester disease (n=1) with BRAF V600E mutation in the tumor tissue. Of these 33 patients, 25 (76%) had the same mutation in urinary cell-free DNA (mutant, n=14; low-mutant, n=11). In addition, 17 patients had longitudinal analysis of percentage of BRAF V600E mutation to wild-type in sequentially collected urine samples and the dynamics of BRAF V600E mutation in the urine correlated with response to therapy in 13 (76%) of them.

Conclusion: Our data suggest that detecting BRAF V600E mutation in cell-free DNA from urine can offer a noninvasive alternative to mutation testing of tumor tissue with acceptable concordance and should be investigated further for testing and monitoring of mutation status in patients with cancer and Erdheim-Chester disease.

Citation Format: Filip Janku, Gerald S. Falchook, Sarina A. Piha-Paul, Aung Naing, Veronica R. Holley, David S. Hong, Vivek Subbiah, Daniel D. Karp, Ralph G. Zinner, Siqing Fu, Jennifer J. Wheler, Funda Meric-Bernstam, Vanda M. Stepanek, Rajyalakshmi Luthra, Lorieta Leppin, Latifa Hassaine, Karena Kosco, Jason C. Poole, Cecile Rose T. Vibat, Mark G. Erlander. Droplet digital PCR detection and longitudinal monitoring of BRAF mutations in cell-free urinary DNA of patients with metastatic cancers or Erdheim-Chester disease. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr LB-170. doi:10.1158/1538-7445.AM2014-LB-170

Aberrant recombination involving the granzyme locus occurs in Atm-/- T-cell lymphomas

Ataxia telangiectasia (A-T) is an autosomal recessive disease caused by loss of function of the serine/threonine protein kinase ATM (ataxia telangiectasia mutated). A-T patients have a 250-700-fold increased risk of developing lymphomas and leukemias which are typically highly invasive and proliferative. In addition, a subset of adult acute lymphoblastic leukemias and aggressive B-cell chronic lymphocytic leukemias that occur in the general population show loss of heterozygosity for ATM. To define the specific role of ATM in lymphomagenesis, we studied T-cell lymphomas isolated from mice with mutations in ATM and/or p53 using cytogenetic analysis and mRNA transcriptional profiling. The analyses identified genes misregulated as a consequence of the amplifications, deletions and translocation events arising as a result of ATM loss. A specific recurrent disruption of the granzyme gene family locus was identified resulting in an aberrant granzyme B/C fusion product. The combined application of cytogenetic and gene expression approaches identified specific loci and genes that define the pathway of initiation and progression of lymphoreticular malignancies in the absence of ATM.