Molecular Tests for Risk-Stratifying Cytologically Indeterminate Thyroid Nodules: An Overview of Commercially Available Testing Platforms in the United States

The past decade has witnessed significant advances in the application of molecular diagnostics for the pre-operative risk-stratification of cytologically indeterminate thyroid nodules. The tests that are currently marketed in the United States for this purpose combine aspects of tumor genotyping with gene and/or microRNA expression profiling. This review compares the general methodology and clinical validation studies for the three tests currently offered in the United States: ThyroSeq v3, Afirma GSC and Xpression Atlas, and ThyGeNEXT/ThyraMIR.

Synergizing microdissection with ddPCR to advance precision oncology.

e15083 Background: Improving patient treatment outcomes based on the application of advanced molecular profiling methodologies applied to pathological specimens is the purview of precision oncology. Optimal utilization of tumor tissue from diagnostic biopsies remains an unmet medical need. This is especially relevant today since precision oncology is a rapidly evolving field where timely tumor genotyping is essential for the indication of many advanced and targeted therapies. National Comprehensive Cancer Network (NCCN) guidelines now mandate molecular testing for clinically actionable targets in certain malignancies. Patients diagnosed with advanced non-small cell lung cancer (NSCLC) are commonly of an older age and have significant co-morbidities. Clinical dilemmas regarding the ability to obtain adequate amounts of tissue for tumor genotyping frequently occur. In these cases, the tumor tissue may have been obtained by an image-guided biopsy, and the diagnosis of NSCLC proper determined via cytology. Adequate tissue for tumor genotyping and/or a more advanced mutational analysis to identify oncogenic drivers may not be available, and the opportunity to explore more optimal therapy options (eg, TKI in lieu of chemotherapy) is lost. Methods: Following a formal pathology review formalin fixed paraffin embedded (FFPE) specimens were examined using advanced immuno-based laser capture microdissection (LCM). In preparation for droplet digital PCR (ddPCR), nucleic acids were extracted from samples and run with a series of positive and negative controls. Results: Utilizing lung cancer as an example, an improved genotyping approach for NSCLC solid tumors was developed and tested. The strategy involves optimization of the microdissection process and analysis of a large number of identical target cells from FFPE specimens sharing similar characteristics, in other words, single-cell subtype analysis. Immunostaining status, cell phenotype, and spatial location within a histological section are examples of shared characteristics that can guide cell procurement. Conclusions: With this approach, tumor molecular analysis is enhanced through the synergy between microdissection and ddPCR. Here we demonstrate a methodology that illustrates genotyping of a solid tumor from a small tissue biopsy sample in a time and cost-efficient manner, using immunohistochemistry-directed LCM along with the detection and absolute quantitation of mutations by ddPCR.

Utilizing single-cell-subtype analysis via advanced microdissection methods and the impact on precision medicine.

e15571 Background: Improving the utilization of tumor tissue from diagnostic biopsies is an unmet medical need. This is especially relevant today in the rapidly evolving precision oncology field where tumor genotyping is essential for the indication of many advanced and targeted therapies. National Comprehensive Cancer Network (NCCN) guidelines now mandate molecular testing for clinically actionable targets in certain malignancies. Patients diagnosed with advanced non-small cell lung cancer (NSCLC) are commonly of an older age and have significant co-morbidities. This frequently causes clinical dilemmas regarding the ability to obtain adequate amounts of tissue for tumor genotyping. In these cases, the tumor tissue may have been obtained by an image-guided biopsy, and the diagnosis of NSCLC proper determined via cytology. In certain instances, adequate tissue for tumor genotyping and/or a more advanced mutational analysis to identify oncogenic drivers may not be available. Methods: Following pathology review, formalin fixed paraffin embedded (FFPE) specimens were subjected to advanced immuno-based laser capture microdissection (LCM), DNA was extracted and prepped for droplet digital PCR (ddPCR) and run with a series of positive and negative controls. Results: Utilizing advanced stage lung cancer as an example, an improved genotyping approach for solid tumors is possible. The strategy involves optimization of the microdissection process and analysis of a large number of identical target cells from FFPE specimens sharing similar characteristics, in other words, single-cell subtype analysis. The shared characteristics can include immunostaining status, cell phenotype, and/or spatial location within a histological section. Conclusions: Synergy between microdissection and ddPCR enhances the molecular analysis. We demonstrate a methodology that illustrates genotyping of a solid tumor from a small tissue biopsy sample in a time and cost-efficient manner, using immunostain targeting as an example.

Genomic biomarkers to determine survival in Multicenter Study of RAS mutations (MURAS) in patients with colorectal liver metastases receiving Y90 radioembolization treatment.

e16119 Background: Assessment of the clinical outcomes and prognostic value of genomic mutations in colorectal liver metastases treated with Y90 radioembolization. Methods: Multi-institution retrospective study of patients who underwent Y-90 radioembolization treatment after tumor genotyping was completed for CRLM. Patients treated between 2008 and 2019 were included from 7 institutions within the United States and Europe. Patient demographics, tumor characteristics, pre- and post- treatment regimens, serum laboratory evaluation and overall survival were analyzed between patients with differing histopathologic and genomic status. Tumor genotyping was obtained for KRAS, BRAF, PIK3CA, AKT, MEK, NRAS and MMR genes. Kaplan-Meier survival estimation and multivariate Cox regression were analyzed. Results: 434 patients treated with Y90 radioembolization fulfilled the inclusion criteria. Of the total cohort, 399 patients were available who had sufficient documented tumor profiling data. Average age at diagnosis was found to be 58.8 years for all patients (60.1% male, 39.9% female). Decreased survival post Y-90 treatment was shown in those patients with increased number of documented tumor mutations (n = 0,1 or ≥2 mutations: median OS 9.63 mos vs. 6.2 mos vs. 5.3 mos; p < 0.0001). Additionally, the median survival in patients with mutated BRAF was 5.0 months, as compared to 9.4 months in those patients with wild-type BRAF (p = 0.0009). Primary colon cancer sidedness was also shown to demonstrate significant difference in survival post-Y90 treatment with left sided primaries showing improved median overall survival (left = 7.5 mos vs. right = 6.3 mos; p = 0.04.) Patient demographics including gender, age and race were not shown to be significant in overall survival post-Y90 treatment (p-values > 0.05). Number of tumor mutations (p < 0.0001, HR = 1.69 CI: 1.39-2.05), BRAF status (p = 0.02, HR = 2.6 CI:1.20-4.9), primary sidedness (p = 0.01, HR = 0.65 CI:0.47-0.90), pre-treatment neutrophil-lymphocyte ratio (p = 0.04, HR = 1.42 CI: 1.02-1.98) and KRAS mutation status (p < 0.0001, HR = 1.81 CI: 1.45-2.26) all persisted as significant predictors of survival on multivariate analysis. Conclusions: Number of tumor mutations, BRAF mutation status, primary tumor sidedness, neutrophil-lymphocyte ratio and KRAS mutation status are all shown to be significant prognostic factors in patients with colorectal liver metastases receiving Y90 radioembolization.

KEAP1/NFE2L2 mutations to predict local recurrence after radiotherapy but not surgery in localized non-small cell lung cancer.

9012 Background: Tumor genotyping in localized non-small cell lung cancer (NSCLC) is not broadly performed due to lack of actionable associations of mutations with treatment or outcome. We sought to identify recurrent mutations in localized NSCLC that are associated with local recurrence (LR) after radiotherapy (RT) or surgery. Methods: We identified consecutive patients with NSCLC treated with chemoradiotherapy (CRT), stereotactic ablative radiotherapy (SABR) or surgery from 2009-2018 at our institution with stage IA1-IIIC NSCLC who had genotyping performed on tumor tissue using a targeted gene panel. Our primary objective was to identify somatic tumor mutations that predicted LR after RT but not surgery. We also performed functional screening assays by expressing open reading frame constructs harboring patient-derived mutations in knock-out cell lines generated by CRISPR-Cas9 and evaluating effects on in vitro radioresistance. Results: We identified 232 consecutive patients with localized NSCLC (87.1% adenocarcinoma, 10.3% squamous, 2.3% other) who received tumor biopsy or resection specimen and underwent tumor genotyping. 47 patients with locally advanced NSCLC received CRT, 50 patients with early stage NSCLC received SABR, and 135 patients with early stage NSCLC underwent surgical resection. Of all recurrent mutations ( > 5% mutation frequency), only mutations in Kelch-like ECH-associated protein 1 ( KEAP1) or Nuclear Factor Erythroid 2-Related Factor 2 ( NFE2L2) genes (K/NMUT) were significantly associated with LR after CRT or SABR, with 2-year LR in the combined RT cohort of 42.4% for K/NMUT versus 12.5% for wildtype (P = 0.005). Furthermore, K/NMUT were present in nearly half of all LR events. Strikingly, there was no significant difference in LR for K/NMUT tumors following CRT versus SABR (P = 0.47). Local recurrence was rare for patients who received surgery (n = 2) and was not associated with K/N mutation status (P = 0.60). Functional evaluation by expression of K/N mutations in knock-out cell lines revealed that LR only occurred in patients with mutations that induced radioresistance (i.e. pathogenic) but not passenger mutations (P = 0.04). In contrast to genotyping, NFE2L2 target gene expression analysis via RNA-seq did not predict LR (P = 0.93). Conclusions: Our findings suggest that KEAP1/NFE2L2 mutations are a predictive biomarker of clinical radioresistance and a dominant cause of LR after RT. Genotyping for KEAP1/NFE2L2 mutations could therefore facilitate treatment personalization in localized NSCLC.

Feasibility and Benefit of Molecularly-Informed Enrollment into Personalized Therapies or Early Phase Trials for Patients with Relapsed or Refractory Multiple Myeloma

Background: Treatments of patients with relapsed or refractory multiple myeloma (r/r MM) remains a challenge and there is no molecular-informed personalized therapies available in this context. Cytogenetics and next-generation sequencing (NGS) panels can rapidly identify recurrent molecular abnormalities, thus helping to orient patients (pts) in appropriate targeted therapies or clinical trials. We aimed to evaluate whether selecting pts through tumor genotyping is associated with a better outcome. Methods: From 2013 to 2018, all pts with r/r MM screened for molecular and/or cytogenetics before enrollment in early clinical trials (eaCTs) were included. Molecular screening methods included bone marrow cytogenetics, sanger assays for BRAF screening or next-generation sequencing on sorted CD138 positive bone-marrow cells. The actionable targets and therapies related were BRAF V600E mutation with BRAF inhibitor and t(11;14) with BCL2 inhibitor. The objective were to evaluate the feasibility and potential benefit of using tumor genotyping to orient patients with molecularly-informed multiple myeloma in personalized therapies or eaCTs. The tumor responses rates, median duration of treatments and overall survival (OS) were assessed in molecularly oriented (MO) and non-molecularly oriented (non-MO) pts. Efficacy was evaluated using International Myeloma Working Group Uniform Response Criteria. Results: Forty-six pts with r/r MM were enrolled, mean age was 66 y (range 52-81), median of previous lines of therapies was 3 (range 1-8). Prior systemic therapies included immunomodulatory agents (n=46, 100%), alkylating agents (n=43, 94%) or proteasome inhibitors (n=44, 95%), and 28 (62%) pts had previously received auto stem-cell transplant. Identification of potentially actionable targets was found in 13 (28%) pts, including 8 (17%) pts with t(11;14) and 5 (11%) pts with BRAF V600E mutations. Eight (17%) out of the 46 pts were treated in molecularly oriented (MO) personalized therapies or eaCTs, and 38 (83%) pts were treated in non-MO therapies or eaCTs. The MO pts received BRAF inhibitor alone or in combination with MEK inhibitor (n=5), or BCL2 inhibitor given in combination with bortezomib and dexamethasone (n=3). The overall response rate was 75% (4 VGPR, 1 PR and 1 CR) in MO pts versus 11% (4 PR) in non-MO pts (p<0.0001). The median decrease of serum monoclonal component was -94% (range: -99; -55) in MO pts versus -4% (range: -72; +967) in the non-MO pts (p<0.0001). The median durations of treatment were 7.3 [CI95: 0.5-29.0] months and 2.3 [CI95: 1.7-8.0] months in MO and non-MO pts, respectively (p=0.009). The median OS were not reached in MO pts and 43 months in non-MO pts (p=0.76; HR=0.8 [CI95: 0.2-3.4]). Conclusions: Molecular-oriented treatments of relapse or recurrent multiple myeloma can be associated with higher responses rates and prolonged durations on therapy. Accelerating the use of prospective genomics tumor molecular portraits may increase the chances of precision medicine for patients with relapse or recurrent multiple myeloma. Disclosures Le Bras: Amgen: Consultancy. Belhadj:Amgen: Consultancy, Honoraria; Janssen: Consultancy, Honoraria; Takeda: Consultancy, Honoraria; Celgene: Consultancy, Honoraria. Soria:Medimmune: Employment. Ribrag:Amgen: Research Funding; Gilead: Consultancy, Honoraria; Infinity: Consultancy, Honoraria; Incyte Corporation: Consultancy; BMS: Consultancy, Honoraria, Other: travel; epizyme: Consultancy, Honoraria; argenX: Research Funding; MSD: Honoraria; pharmamar: Other: travel; Servier: Consultancy, Honoraria; NanoString Technologies: Consultancy, Honoraria; Roche: Honoraria, Other: travel.