The NCI-MATCH trial: lessons for precision oncology

The NCI-MATCH (Molecular Analysis for Therapy Choice) trial ( NCT02465060 ) was launched in 2015 as a genomically driven, signal-seeking precision medicine platform trial-largely for patients with treatment-refractory, malignant solid tumors. Having completed in 2023, it remains one of the largest tumor-agnostic, precision oncology trials undertaken to date. Nearly 6,000 patients underwent screening and molecular testing, with a total of 1,593 patients (inclusive of continued accrual from standard next-generation sequencing) being assigned to one of 38 substudies. Each substudy was a phase 2 trial of a therapy matched to a genomic alteration, with a primary endpoint of objective tumor response by RECIST criteria. In this Perspective, we summarize the outcomes of the initial 27 substudies in NCI-MATCH, which met its signal-seeking objective with 7/27 positive substudies (25.9%). We discuss key aspects of the design and operational conduct of the trial, highlighting important lessons for future precision medicine studies.

Recommendations for the Use of in Silico Approaches for Next-Generation Sequencing Bioinformatic Pipeline Validation: A Joint Report of the Association for Molecular Pathology, Association for Pathology Informatics, and College of American Pathologists

In silico approaches for next-generation sequencing (NGS) data modeling have utility in the clinical laboratory as a tool for clinical assay validation. In silico NGS data can take a variety of forms, including pure simulated data or manipulated data files in which variants are inserted into existing data files. In silico data enable simulation of a range of variants that may be difficult to obtain from a single physical sample. Such data allow laboratories to more accurately test the performance of clinical bioinformatics pipelines without sequencing additional cases. For example, clinical laboratories may use in silico data to simulate low variant allele fraction variants to test the analytical sensitivity of variant calling software or simulate a range of insertion/deletion sizes to determine the performance of insertion/deletion calling software. In this article, the Working Group reviews the different types of in silico data with their strengths and limitations, methods to generate in silico data, and how data can be used in the clinical molecular diagnostic laboratory. Survey data indicate how in silico NGS data are currently being used. Finally, potential applications for which in silico data may become useful in the future are presented.

Impact of Somatic Mutations on Survival Outcomes in Patients With Anaplastic Thyroid Carcinoma

PURPOSE

Anaplastic thyroid carcinoma (ATC) uniformly present with aggressive disease, but the mutational landscape of tumors varies. We aimed to determine whether tumor mutations affect survival outcomes in ATC.

MATERIALS AND METHODS

Patients who underwent mutation sequencing using targeted gene panels between 2005 and 2019 at a tertiary referral center were included. Associations between mutation status and survival outcomes were assessed using Cox proportional hazards models.

RESULTS

A total of 202 patients were included, where 122 died of ATC (60%). The median follow-up was 31 months (interquartile range, 18-45 months). The most common mutations were in TP53 (59%), BRAF (41%), TERT promoter (37%), and the RAS gene family (22%). Clinicopathologic characteristics and overall survival (OS) significantly correlated with mutations in BRAFV600E and RAS, which were mutually exclusive. The BRAFV600E mutation was associated with the presence of a papillary thyroid carcinoma precursor and significantly better OS (median OS: 24 months). RAS-mutated patients more commonly presented without cervical lymph node involvement but had the worst OS (median OS: 6 months). Tumors that were wild-type for both BRAF and RAS were enriched for NF1 mutations and harbored intermediate prognosis (median OS: 15 months). In multivariate analyses, RAS mutations were associated with a more than 2.5-fold higher risk of death (adjusted hazard ratio, 2.64; 95% CI, 1.66 to 4.20) compared with BRAFV600E. In patients treated with BRAF-directed therapy (n = 60), disease progression occurred in 48% of patients (n = 29). The median progression-free survival was 14 months. The presence of a TP53 mutation was independently associated with reduced progression-free survival in BRAFV600E-mutated patients treated with BRAF-directed therapy (adjusted hazard ratio, 2.89; 95% CI, 1.35 to 6.21).

CONCLUSION

Mutation analysis provides prognostic information in ATC and should be incorporated into routine clinical care.

Poziotinib for EGFR exon 20-mutant NSCLC: Clinical efficacy, resistance mechanisms, and impact of insertion location on drug sensitivity

We report a phase II study of 50 advanced non-small cell lung cancer (NSCLC) patients with point mutations or insertions in EGFR exon 20 treated with poziotinib (NCT03066206). The study achieved its primary endpoint, with confirmed objective response rates (ORRs) of 32% and 31% by investigator and blinded independent review, respectively, with a median progression-free survival of 5.5 months. Using preclinical studies, in silico modeling, and molecular dynamics simulations, we found that poziotinib sensitivity was highly dependent on the insertion location, with near-loop insertions (amino acids A767 to P772) being more sensitive than far-loop insertions, an observation confirmed clinically with ORRs of 46% and 0% observed in near versus far-loop, respectively (p = 0.0015). Putative mechanisms of acquired resistance included EGFR T790M, MET amplifications, and epithelial-to-mesenchymal transition (EMT). Our data demonstrate that poziotinib is active in EGFR exon 20-mutant NSCLC, although this activity is influenced by insertion location.

Actionable Tumor Alterations and Treatment Protocol Enrollment of Pediatric and Young Adult Patients With Refractory Cancers in the National Cancer Institute-Children's Oncology Group Pediatric MATCH Trial

PURPOSE

The National Cancer Institute-Children's Oncology Group Pediatric MATCH trial aimed to facilitate evaluation of molecular-targeted therapies in biomarker-selected cohorts of childhood and young adult patients with cancer by screening tumors for actionable alterations.

PATIENTS AND METHODS

Tumors from patients age 1-21 years with refractory solid tumors, lymphomas, or histiocytic disorders were subjected to cancer gene panel sequencing and limited immunohistochemistry to identify actionable alterations for assignment to phase II treatment arms. The rates of treatment arm assignment and enrollment were compared between clinical and demographic groups.

RESULTS

Testing was completed for 94.7% of tumors submitted. Actionable alterations were detected in 31.5% of the first 1,000 tumors screened, with treatment arm assignment and enrollment occurring in 28.4% and 13.1% of patients, respectively. Assignment rates varied by tumor histology and were higher for patients with CNS tumors or enrolled at Pediatric Early Phase Clinical Trials Network sites. A reported history of prior clinical molecular testing was associated with higher assignment and enrollment rates. Actionable alterations in the mitogen-activated protein kinase signaling pathway were most frequent (11.2%). The most common reasons provided for not enrolling on treatment arms were patients receiving other treatment or poor clinical status.

CONCLUSION

The Pediatric MATCH trial has proven the feasibility of a nationwide screening Protocol for identification of actionable genetic alterations and assignment of pediatric and young adult patients with refractory cancers to trials of molecularly targeted therapies. These data support the early use of tumor molecular screening for childhood patients with cancer whose tumors have not responded to standard treatments.

Only SF3B1 mutation involving K700E independently predicts overall survival in myelodysplastic syndromes

BACKGROUND

SF3B1 mutations (SF3B1^mut ) in myelodysplastic syndromes (MDS) frequently involve codon K700E and have a favorable prognosis. The prognostic effect of non-K700E SF3B1^mut is uncertain.

METHODS

The authors analyzed the clinicopathological features and outcomes of a single-institution series of 94 treatment-naive SF3B1^mut MDS patients (18%) and 415 treatment-naive SF3B1^wt MDS patients and explored the differences between K700E and non-K700E SF3B1^mut MDS.

RESULTS

Fifty-five patients (59%) carried K700E. Recurrent non-K700E mutations (39 [41%]) included R625, H662, and K666. Compared with SF3B1^mut K700E patients, non-K700E patients had a lower median absolute neutrophil count (1.8 vs 2.4; P = .005) and were frequently "high" according to the Revised International Prognostic Scoring System (19% vs 4%; P = .031). Non-K700E MDS was associated frequently with RUNX1 (26% vs 7%; P = .012) and exclusively with BCOR, IDH2, and SRSF2 mutations. A splicing analysis showed the differential distribution of alternatively spliced events and gene expression profiles between K700 and non-K700E MDS patients. The majority (at least 80%) of SF3B1^mut K700E, SF3B1^mut non-K700E, and SF3B1^wt patients were treated with hypomethylating agents. Over a median follow-up of 16 months, SF3B1^mut had superior overall survival (OS) in comparison with SF3B1^wt in all MDS patients (not reached vs 25.2 months; P = .0003), in patients with low-grade MDS, and in patients with myelodysplastic syndromes with ring sideroblasts (MDS-RS). Compared with SF3B1^wt , SF3B1^mut K700E had superior outcomes in all MDS (median OS, 25 months vs not reached; P = .0001), in low-grade MDS (median OS, 41.3 months vs not reached; P = .0015), and in MDS-RS (median OS, 22.3 months vs not reached; P = .0001), but no significant difference was seen between non-K700E and SF3B1^wt MDS. By multivariable analysis, the absence of SF3B1^mut K700E mutations was independently associated with the prognosis.

CONCLUSIONS

This study highlights the importance of the SF3B1 mutation subtype in MDS risk assessment.

LAY SUMMARY

Myelodysplastic syndromes (MDS) with SF3B1 mutations are regarded as having a favorable prognosis by both the World Health Organization and the International Working Group for the Prognosis of Myelodysplastic Syndromes. However, this article shows that only MDS patients with SF3B1 K700E mutations have a favorable prognosis (and not MDS patients with SF3B1 mutations involving other codons). This has important implications for refining future MDS subclassification and risk assessment criteria.

Assessment of HER2 (ERBB2) amplification (HER2amp) using blood-based circulating tumor DNA (ctDNA) next generation sequencing (NGS) and correlation with tissue-based testing in metastatic colorectal cancer (mCRC)

3589

Background: HER2 amplified mCRC has emerged as a unique clinical subset, characterized by resistance to anti-EGFR therapy and response to anti-HER2 strategies. Accurate identification and quantification of HER2amp has predictive value for efficacy of anti-HER2 therapies and appropriate patient selection. Despite availability and use of various tumor tissue-based and blood-based assays for detecting HER2amp, data on cross-performance of these platforms are lacking. Methods: Leveraging a multicenter international consortium (Italy, Japan and USA), we generated a large cohort (N = 353) of mCRC patients (pts), tested for HER2amp using both tissue and blood. Tissue testing was done using immunohistochemistry (IHC), in-situ hybridization (ISH) and (NGS). ctDNA NGS was performed using CLIA-certified Guardant360 ctDNA assay, capable of detecting HER2 copy number (CN) variations. The primary endpoint was to correlate HER2 gene CNs in tissue (tCN) and plasma (pCN). Descriptive statistics, spearman correlation (r) and Fisher’s exact test were used. Results: Baseline tumors characteristics included right-sided primary in 234 (23%), proficient mismatch repair in 264 (98%) and RAS/BRAF wild type (WT) genotype in 194 (67%) pts. Tissue testing was done by IHC, ISH and NGS in 76%, 64% and 74% pts, respectively. A total of 177 pts had HER2amp detected by at least one test: 116 (66%), 157 (89%) and 96 (54%) of which had tissue +, ctDNA +, and both tissue and ctDNA + disease, respectively. Discordant cases consisted of 20 (6%) with positivity in tumor only and 61 (17%) in ctDNA only. Sensitivity, specificity, positive and negative predictive values of ctDNA assay (vis-à-vis tissue) were 83%, 74%, 61% and 90% respectively. Among HER2amp pts, median (range) HER2/CEP17 (ISH) ratio, tCN and pCN were 5.2 (2–12), 11.6 (2–700) and 3.5 (2–122), respectively. The pCN showed strong correlation with ISH ratio (r = 0.69) and tCN (r = 0.68) (P < 0.001). Median pCN differed significantly between pts with HER2 IHC 3+ (12.0), 2+ (2.2) and 0/1+ (2.0) tumors (P < 0.001). High HER2amp (pCN > 4.0) appeared to be enriched with tissue + cases (69% vs 8% [OR 24.6, P < 0.001]), tumor tissue HER2 + status (IHC3+ [75%] vs IHC2+ISH+ [50%] vs IHC2+/ISH- or IHC0/1+ [12%], P < 0.001), HER2 tCN > 6 (79% vs 31% [OR 8.7, P < 0.001]) and RAS/BRAF WT tumors (41% vs 17% [OR 3.5, P = 0.064) but not left sidedness (41% vs 38%; OR 1.1; P = 0.82). Conclusions: In this large diverse cohort of mCRC, we demonstrated correlation of HER2 tCN and pCN obtained by tissue-based and blood-based ctDNA assay. Further prospective efforts are needed to standardize this cross-platform quantification of HER2amp to facilitate robust clinical application of HER2 therapies. This effort shows the value of strategic international partnership in furthering research for rare cancer subsets.

Factors impacting enrollment on NCI-COG Pediatric MATCH trial treatment protocols

10007

Background: The NCI-Children’s Oncology Group (COG) Pediatric Molecular Analysis for Therapy Choice (MATCH) trial assigns patients age 1 to 21 years with relapsed or refractory solid tumors, lymphomas, and histiocytic disorders to phase 2 treatment arms of molecularly-targeted therapies based on the genetic alterations detected in their tumor. Treatment arm assignments and enrollment decisions have now been made for 1000 study participants: we report here match and enrollment data and factors affecting treatment protocol enrollment. Methods: Patients enrolled in the Pediatric MATCH screening protocol were assigned to an open treatment protocol if an actionable mutation (aMOI) was detected by tumor DNA and RNA-based cancer gene panel sequencing. After a match, treatment protocol enrollment must occur within 8-12 weeks. Patient demographic data, reasons for not enrolling on treatment protocol (if applicable), and prior history of molecular testing were reported by study sites. The Fisher exact test was used to compare protocol enrollment rates between groups. Results: Results were analyzed for the first 1000 patients with testing completed (enrolled between July 2017 and October 2020). At least one tumor aMOI was detected in 310 (31%) patients and treatment protocol slots were available for 284 patients (28%). A total of 131 patients (46% of those matched) enrolled on a treatment arm. No difference in treatment protocol match or enrollment rate was observed for gender, race, or ethnicity. Both treatment protocol match rate (105/275, 38% vs 86/394, 22%) and enrollment rate (56/275, 20% vs 33/394, 8%) were significantly more frequent in patients with a reported history of prior molecular testing (p<0.0001). The most common reasons provided for not enrolling on a treatment protocol were: patient receiving other treatment (32% of responses), poor clinical status (16%), lack of measurable disease (11%), or ineligible diagnosis for that treatment arm (10%). Ineligibility due to history of excluded prior targeted therapy (6%) or inability to swallow capsules (4%) was less frequent. Conclusions: The rate of Pediatric MATCH treatment protocol enrollment has exceeded pre-study projections, due to more frequent actionable mutation detection and treatment assignment than anticipated (28% observed, 10% projected). This may in part reflect an increased number of targetable events in recurrent or refractory pediatric cancers. Correlative studies analyzing pre-treatment tumors from MATCH study patients are underway and will address this hypothesis. Prior history of molecular testing was associated with higher match and enrollment rate and poor clinical status was a common reason for not enrolling on a treatment protocol, suggesting that early molecular screening of children with solid malignancies may facilitate enrollment to biomarker-selected trials of targeted therapies. Clinical trial information: NCT03155620.

The Molecular Analysis for Therapy Choice (NCI-MATCH) Trial: Lessons for Genomic Trial Design

Background

The proportion of tumors of various histologies that may respond to drugs targeted to molecular alterations is unknown. NCI-MATCH, a collaboration between ECOG-ACRIN Cancer Research Group and the National Cancer Institute, was initiated to find efficacy signals by matching patients with refractory malignancies to treatment targeted to potential tumor molecular drivers regardless of cancer histology.

Methods

Trial development required assumptions about molecular target prevalence, accrual rates, treatment eligibility, and enrollment rates as well as consideration of logistical requirements. Central tumor profiling was performed with an investigational next-generation DNA–targeted sequencing assay of alterations in 143 genes, and protein expression of protein expression of phosphatase and tensin homolog, mutL homolog 1, mutS homolog 2, and RB transcriptional corepressor 1. Treatments were allocated with a validated computational platform (MATCHBOX). A preplanned interim analysis evaluated assumptions and feasibility in this novel trial.

Results

At interim analysis, accrual was robust, tumor biopsies were safe (<1% severe events), and profiling success was 87.3%. Actionable molecular alteration frequency met expectations, but assignment and enrollment lagged due to histology exclusions and mismatch of resources to demand. To address this lag, we revised estimates of mutation frequencies, increased screening sample size, added treatments, and improved assay throughput and efficiency (93.9% completion and 14-day turnaround).

Conclusions

The experiences in the design and implementation of the NCI-MATCH trial suggest that profiling from fresh tumor biopsies and assigning treatment can be performed efficiently in a large national network trial. The success of such trials necessitates a broad screening approach and many treatment options easily accessible to patients.

Molecular Landscape and Actionable Alterations in a Genomically Guided Cancer Clinical Trial: National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH)

Therapeutically actionable molecular alterations are widely distributed across cancer types. The National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH) trial was designed to evaluate targeted therapy antitumor activity in underexplored cancer types. Tumor biopsy specimens were analyzed centrally with next-generation sequencing (NGS) in a master screening protocol. Patients with a tumor molecular alteration addressed by a targeted treatment lacking established efficacy in that tumor type were assigned to 1 of 30 treatments in parallel, single-arm, phase II subprotocols.

A Cryptic BCR-PDGFRB Fusion Resulting in a Chronic Myeloid Neoplasm With Monocytosis and Eosinophilia: A Novel Finding With Treatment Implications

RNA-seq was used to identify the partner gene and confirm the presence of a BCR-PDGFRB fusion. Identification of this fusion product resulted in successful treatment and long-term remission of this myeloid neoplasm. Based on our results, we suggest that despite current WHO recommendations, screening for PDGFRB rearrangement in cases of leukocytosis with eosinophilia and no other etiologic explanation is necessary, even if the karyotype is normal.

Germline DNMT3A mutation in familial acute myeloid leukaemia

Acute myeloid leukaemia (AML) is a heterogeneous myeloid malignancy characterized by recurrent clonal events, including mutations in epigenetically relevant genes such as DNMT3A, ASXL1, IDH1/2, and TET2. Next-generation sequencing analysis of a mother and son pair who both developed adult-onset diploid AML identified a novel germline missense mutation DNMT3A p.P709S. The p.P709S protein-altering variant resides in the highly conserved catalytic DNMT3A methyltransferase domain. Functional studies demonstrate that the p.P709S variant confers dominant negative effects when interacting with wildtype DNMT3A. LINE-1 pyrosequencing and reduced representation bisulphite sequencing (RBBS) analysis demonstrated global DNA hypomethylation in germline samples, not present in the leukaemic samples. Somatic acquisition of IDH2 p.R172K mutations, in concert with additional acquired clonal DNMT3A events in both patients at the time of AML diagnosis, confirms the important pathogenic interaction of epigenetically active genes, and implies a strong selection and regulation of methylation in leukaemogenesis. Improved characterization of germline mutations may enable us to better predict malignant clonal evolution, improving our ability to provide customized treatment or future preventative strategies.

Outcomes of acute myeloid leukemia with myelodysplasia related changes depend on diagnostic criteria and therapy

Acute myeloid leukemia with myelodysplasia-related changes (AML-MRC) is a heterogeneous disorder defined by multilineage dysplasia, myelodysplastic syndrome (MDS)-related karyotype, or history of prior MDS. We evaluated 415 patients with AML-MRC treated from 2013 to 2018 and analyzed their clinical outcomes based on the diagnostic criteria of AML-MRC, therapy type and mutation profile. Criteria for AML-MRC included: cytogenetic abnormalities (AML-MRC-C) in 243 (59%), prior history of MDS in 75 (18%) including 47 (11%) with previously untreated MDS (AML-MRC-H) and 28 (7%) with previously treated MDS (AML-MRC-TS), and 97 (23%) with multilineage dysplasia (AML-MRC-M). Median age was 70 years (range 18-94). Among 95 evaluable patients, a total of 37 (39%) had secondary-type (ASXL1, BCOR, EZH2, SF3B1, SRSF2, STAG2, U2AF1, ZRSR2) mutations. Mutations in ASXL1, BCOR, SF3B1, SRSF2, and U2AF1 tended to appear in dominant clones. By multivariate analysis, AML-MRC subtype, age and serum LDH levels were independent predictors of outcome, with patients with AML-MRC-M (HR 0.56, CI 0.38-0.84, P = .004) and AML-MRC-H having better OS. Compared to a cohort of 468 patients with AML without MRC, patients with AML-MRC-M/AML-MRC-H had similar outcomes to those with intermediate risk AML by European LeukemiaNet criteria. Intensive therapy was associated with improved OS in patients with AML-MRC-M (HR 0.42, CI 0.19-0.94, P = .036) and with improved EFS in AML-MRC-M and AML-MRC-H (HR 0.26, CI 0.10-0.63, P = .003). This data suggests that not all diagnostic criteria for AML-MRC define high-risk patients and that specific subgroups may benefit from different therapeutic interventions.

Performance of cell-free tumor DNA testing for 101 clinical laboratories on College of American Pathologists proficiency tests

e13681

Background: Cell-free tumor DNA or circulating tumor DNA tests are increasingly used in clinical care to detect somatic mutations from solid tumors. However, data on laboratory performance characteristics using standardized samples is limited. Methods: Well-characterized reference materials were used for the College of American Pathologists (CAP) cell-free tumor DNA proficiency testing surveys, which consisted of stabilized DNA fragmented to simulate cell-free DNA in a synthetic plasma matrix. For the 2018A, 2018B, 2019A and 2019B surveys, laboratories tested for hotspot mutations (single and dinucleotide sequence changes) in EGFR, BRAF, KRAS, NRAS, and IDH1 at variant allele fractions ranging from 0.1% - 1.0%. As per CAP proficiency testing standards, results were scored according to the known mutation(s) engineered at designated variant allele fractions in each PT sample. Nine laboratories were excluded from analysis because they provided incomplete results. Statistical significance was calculated using a multivariate logistic regression model. Results: In 2018 and 2019, 101 laboratories submitted survey results for at least one proficiency testing mailing. There were 5088 total proficiency testing responses for EGFR, BRAF, KRAS, NRAS, and IDH1 mutations across 12 different samples. For the 3585 responses submitted for BRAF, KRAS, NRAS, and IDH1, sensitivity ranged from 94.6 – 100%, while specificity exceeded 99%. There were no significant differences in performance between analytical methodologies for BRAF, KRAS, NRAS, and IDH1 mutations. Performance characteristics for EGFR mutations among the 1503 responses showed a combined sensitivity of 87.1% and specificity of 98.7%. For laboratories detecting mutations in EGFR, next-generation sequencing methods exhibited a sensitivity (true positivity) of 95.7% while the sensitivity of non-NGS methods was lower at 81.7% ( P= 0.02). Conclusions: These findings demonstrate high sensitivity and specificity for clinical laboratories performing cell-free tumor DNA tests. For EGFR mutations, NGS outperformed non-NGS methods. These data suggest excellent overall agreement among laboratories performing clinical cell-free tumor DNA tests. Further investigation across variant allele fractions and additional variant types is warranted.

Clinical outcome for gastrointestinal cancers with polymerase epsilon mutations treated with immunotherapy

828

Background: Predictive factors for immunotherapy benefit across gastrointestinal (GI) cancers are limited to those with deficient mismatch repair (dMMR). Mutations in DNA polymerase E (POLE) can cause a hypermutated phenotype irrespective of MMR status and may serve as a predictor factor for immunotherapy. Methods: All POLE-mutant GI cancers with tumor mutation burden (TMB) measured at MD Anderson Cancer Center were identified (N = 58) from 2014 to 2019 and those that had received immunotherapy (N = 18) were studied. TMB was measured by genomic profiling assays FoundationOne or STGA 2018. Samples were analyzed using Kaplan-Meier method and compared using log rank test, with a predefined high TMB threshold of 20 muts/mb. Results: Of the 58 POLE-mutant GI tumors a high TMB was present in 50% of the cases. 18 patients (median age 64, female gender 16.7%, median prior lines of therapy 2) with locally advanced (N = 4) or metastatic (N = 14) POLE-mutant GI cancers (13 colorectal, 1 small intestine, 1 esophageal, 1 pancreatic, and 2 gallbladder) were identified. Immunotherapy agents were PD-1/PD-L1 based in 17 and CTLA-4 based in 1. High TMB was present in 11 patients. dMMR was present in 7 patients, all with high TMB. Treatment response was seen in 3 patients (2 CR and 1 PR), all TMB high and dMMR. Progression-free survival (PFS) was significantly longer in those with high TMB (median: 10.1 months vs. 3 months, p = 0.048). Of the 11 pMMR patients median PFS was 9.3 months in high TMB and 3 months in low TMB (p = 0.246). Conclusions: POLE mutations in GI cancers were associated with a high TMB ( > 20muts/mb) in 50% of patients. Within POLE mutant GI cancers a high TMB may identify patients who benefit from immunotherapy irrespective of mismatch repair status.

Validation of HER2 Amplification as a Predictive Biomarker for Anti–Epidermal Growth Factor Receptor Antibody Therapy in Metastatic Colorectal Cancer

Purpose

HER2 amplification has been implicated in resistance to therapy with anti–epidermal growth factor receptor antibodies (anti-EGFRabs) in metastatic colorectal cancer (mCRC). The purpose of the study was to validate the predictive impact of HER2 amplification in mCRC.

Patients and Methods

We analyzed patients with RAS/BRAF wild-type mCRC across two distinct cohorts. In cohort 1 (n = 98), HER2 amplification was tested in tumor tissue using dual in situ hybridization ( HER2 amplification: HER2/CEP17 ratio, 2.0 or greater). Cohort 2 (n = 70) included 16 patients with HER2 amplification and 54 HER2 nonamplified controls identified by next-generation sequencing ( HER2 amplification: four or more copies) who had received prior anti-EGFRabs. The primary end point was progression-free survival (PFS) on treatment with anti-EGFRab therapy, which was estimated and compared using the Kaplan-Meier method and log-rank test.

Results

Median PFS in cohort 1 on anti-EGFRab–based therapy was significantly shorter in patients with HER2 amplification compared with HER2 nonamplified patients (2.8 v 8.1 months, respectively; hazard ratio [HR], 7.05; 95% CI, 3.4 to 14.9; P < .001). These findings were validated in cohort 2 (median PFS for HER2 amplified v nonamplified: 2.8 v 9.3 months, respectively; HR, 10.66; 95% CI, 4.5 to 25.1; P < .001). The median PFS on therapy without anti-EGFRabs was similar among HER2-amplified and nonamplified patients in both cohort 1 (9.7 v 11.1 months, respectively; HR, 1.01; 95% CI, 0.4 to 2.4; P = .97) and cohort 2 (9.6 v 11.3 months, respectively; HR, 1.21; 95% CI, 0.5 to 3.1; P = .66). In multivariable analyses, HER2 amplification emerged as a single independent predictor of poor PFS on anti-EGFRab therapy in both cohort 1 (HR, 6.48; 95% CI, 3.1 to 13.6; P < .001) and cohort 2 (HR, 10.1; 95% CI, 4.3 to 23.9; P < .001).

Conclusion

HER2 amplification in RAS/RAF wild-type mCRC seems to be a predictive biomarker for lack of efficacy of anti-EGFRab therapy. Screening patients with RAS/BRAF wild-type mCRC for HER2 amplification should be considered before anti-EGFRab treatment to guide therapy and to identify patients for early referral to clinical trials.

Targeting ERBB2 (HER2) Amplification Identified by Next-Generation Sequencing in Patients With Advanced or Metastatic Solid Tumors Beyond Conventional Indications

PURPOSE

Human epidermal growth factor receptor 2 (HER2) is an effective therapeutic target in breast and gastric and gastroesophageal junction cancers. However, less is known about the prevalence of ERBB2 (HER2) amplification and the efficacy of HER2-targeted treatment in other tumors.

PATIENTS AND METHODS

We assessed HER2 amplification status among 5,002 patients with advanced disease (excluding breast cancer) who underwent next-generation sequencing. We evaluated the clinical benefit of HER2-targeted therapy by measuring the time-dependent overall survival (OS) from the genomic testing results, progression-free survival (PFS), and PFS during HER2-targeted therapy (PFS2) compared with PFS during prior therapy (PFS1).

RESULTS

Overall, 122 patients (2.4%) had HER2 amplification, including patients with endometrial (5.3%), bladder (5.2%), biliary or gallbladder (4.9%), salivary (4.7%), and colorectal cancer (3.6%). Forty patients (38%) with nongastric, nongastroesophageal junction, or nonesophageal cancers received at least one line of HER2-targeted therapy. Patients receiving HER2-targeted therapy had a median OS of 18.6 months, compared with 10.9 months for patients who did not receive HER2-targeted therapy (P = .070). On multivariable analysis, HER2-targeted therapy was significantly associated with increased OS (hazard ratio, 0.5; 95% CI, 0.27 to 0.93; P = .029), regardless of sex, age, or number of prior lines of treatment. The PFS2-to-PFS1 ratio was 1.3 or greater in 21 (57%) of 37 patients who received HER2-targeted therapy not in the first line of systemic treatment, and the median PFS2 and PFS1 times were 24 and 13 weeks, respectively (P < .001).

CONCLUSION

HER2 amplifications using next-generation sequencing can be identified in a variety of tumor types. HER2-targeted therapy may confer clinical benefit in tumor types other than those for which HER2 inhibitors are approved.

Pan-Cancer Landscape and Analysis of ERBB2 Mutations Identifies Poziotinib as a Clinically Active Inhibitor and Enhancer of T-DM1 Activity

We characterized the landscape and drug sensitivity of ERBB2 (HER2) mutations in cancers. In 11 datasets (n = 211,726), ERBB2 mutational hotspots varied across 25 tumor types. Common HER2 mutants yielded differential sensitivities to eleven EGFR/HER2 tyrosine kinase inhibitors (TKIs) in vitro, and molecular dynamics simulations revealed that mutants with a reduced drug-binding pocket volume were associated with decreased affinity for larger TKIs. Overall, poziotinib was the most potent HER2 mutant-selective TKI tested. Phase II clinical testing in ERBB2 exon 20-mutant non-small cell lung cancer resulted in a confirmed objective response rate of 42% in the first 12 evaluable patients. In pre-clinical models, poziotinib upregulated HER2 cell-surface expression and potentiated the activity of T-DM1, resulting in complete tumor regression with combination treatment.

Occurrence of other cancers in patients with chronic lymphocytic leukemia and mutations in protection of telomeres 1 (POT1) gene

7529

Background: Mutations in POT1 gene in CLL lead to uncapping of the telomeric ends, causing fusion events and chromosomal aberrations. POT1 is mutated in approximately 4% of pts with CLL. Recent studies reported germline variants in POT1 in pts with familial CLL and in familial melanoma, cardiac angiosarcoma and glioma. We evaluated pts characteristics and the presence of other cancers (OC) in pts with CLL with POT1 mutation seen at our institution. Methods: We performed next generation sequencing (NGS)-based analysis for the detection of mutations in 29 genes frequently mutated in pts with CLL using blood and/or bone marrow samples containing a minimum of 10% clonal B-cells from 1467 pts diagnosed with CLL. Clinical characteristics, prognostic factors (FISH and IGHV status), personal and familial history of OC were collected in pts with POT1 mutations. Results: Mutations in POT1 were found in 52 of the 1467 pts studied (3.5%). Pts with POT1 mutation were young (median age 59 years), commonly presented with early stage disease (Binet stage A 69% vs B/C 31%: p=0.0046 and Rai stage 0-I 65% vs II-IV 35%; p=0.043) and predominantly male (37 male vs 15 female). According to FISH, the more frequent abnormalities were del13q (33%), no abnormalities (25%) and del11q (21%). IGHV status was more commonly unmutated (69%). The most frequent DNA mutations associated with POT1 were NOTCH1 (44%), TP53 (27%) and SF3B1 (23%). Other cancers (excluding non-melanoma skin cancer) were reported in 19 of the 52 pts with POT1 mutation (37%). The most common types were prostate cancer (12%), malignant melanoma (10%) and kidney cancer (8%). Twelve (23%) pts were diagnosed with OC before and 7 pts (13%) after the diagnosis CLL (Table). Four (8%) pts had more than one other cancer diagnoses. Twenty-eight (54%) pts had one of more first degree relatives with history of cancer. Conclusions: POT1 mutations were observed in 3.5% of pts with CLL. These pts are young and often have unmutated IGHV. We observed a high occurrence of OC, particularly malignant melanoma and kidney cancer in these pts. Additional studies are ongoing to determine the proportion of pts with germline POT1 mutation, the distribution of variant allelic frequencies and additional chromosomal abnormalities in pts with OC. Other Cancers in 52 CLL pts with POT1. [Table: see text]

Identification of targetable molecular alterations in the NCI-COG Pediatric MATCH trial

10011

Background: The screening protocol for the NCI-Children’s Oncology Group (COG) Pediatric Molecular Analysis for Therapy Choice (MATCH) trial detects tumor alterations that are used to assign patients with treatment-refractory or recurrent cancers to phase 2 treatment arms of molecularly-targeted therapies. Methods: Patients age 1 to 21 years old with treatment-refractory or recurrent solid tumors, non-Hodgkin lymphomas, or histiocytic disorders treated at U.S. based COG sites are eligible. DNA and RNA extracted from FFPE tumor samples are sequenced using an Oncomine cancer gene panel for detection of mutations, amplifications, and fusions. Loss of SMARCB1, SMARCA4, and PTEN expression is detected by immunohistochemistry. Lists of actionable mutations (aMOIs) based upon available clinical and pre-clinical data are used a priori to determine eligibility for treatment arms. Results: Between 7/24/17 and 12/31/18, 422 patients with a median age of 13 years (range 1-21) were enrolled from 93 COG sites. Solid tumors comprised 71% (n = 300) of diagnoses, CNS tumors 24% (n = 101) and lymphomas/histiocytoses 5% (n = 21). A tumor sample was submitted for 390 patients, sequencing was attempted for 370 (95%), and results were confirmed for 357 (92%). Median turn-around time was 15 days. An aMOI for at least one of the 10 current treatment arms was identified in 112 patients (29%, 95% CI 24%-33%); 95 patients (24%, 95% CI 20%-29%) were assigned to a treatment arm with 39 patients (10%, 95% CI 7%-13%) enrolled to date. The aMOI rate was similar in patients less than 12 years of age (35%) compared to patients 12 years and older (25%). Actionable MAPK pathway alterations were found in 11% of patients (n = 41), most often HRAS/ KRAS/ NRAS mutations (n = 16), BRAF mutations or fusions (n = 14), or NF1 mutations (n = 11). Other genes with recurrent aMOIs included SMARCB1 (n = 14), ALK (n = 8), CDK4 (n = 8), PIK3CA (n = 7), PTEN (n = 7), FGFR1 (n = 5), and BRCA1/BRCA2 (n = 5). Conclusions: Approximately one-quarter of patients with tumor submitted for Pediatric MATCH screening have been assigned to an investigational therapy, facilitating the evaluation of molecularly-targeted agents in biomarker-positive pediatric cohorts through a collaborative nationwide study. Clinical trial information: NCT03155620.

Baseline cfDNA characteristics and evolution of cfDNA profile during treatment with selective FGFR inhibitor TAS-120

3056

Background: There is an increasing role for cfDNA in monitoring response and mechanisms of resistance. We performed cfDNA analysis in a subset of patients enrolled on a Phase I trial with an irreversible, selective FGFR1-4 inhibitor, TAS-120. Methods: 58 plasma samples from 17 patients (13 with cholangiocarcinoma) were analyzed on a 73-gene, next-generation sequencing panel. Selected patients(pts) had longitudinal samples. Results: At least one alteration was detected in 46 cfDNA samples, in 16 (94%) of 17 pts – a pt with GBM had no alterations detected. 14 pts had alterations in FGFR2/3 by genomic testing of archival tumor samples, comprising 20 total alterations (18 unique). 10 of 20 FGFR2/3 alterations were also detected by cfDNA testing: 4/5 SNVs, 1/2 amplifications, 5/13 fusions. Three pts had FGFR/FGF alterations not included (thus not detected) in the cfDNA panel: 2 with FGF ligand amplification, and one FGFR4 mutation. 6 pts (35%) had PR, 5 (29%) had SD and 6 (35%) PD as a best response to TAS-120. Four pts had prior FGFRi: 2 had a PR, 1 SD, and 1 PD on TAS-120. Baseline cfDNA mutations became undetectable during treatment in 4/6 pts with PR. 4 of 6 PD pts had other driver mutations at baseline including mutations in PIK3CA, KRAS, IDH1, BRCA2, or amplifications in PIK3CA, PDGFR. 9 pts with cfDNA available at progression after SD/PR: 3 had acquired FGFR2 mutations (one each of V564L, V564F, or N549K). Two also acquired alterations in other candidate resistance genes ( PTEN and MAP2K1). Another pt had low variant allele frequency (VAF) NRAS G12D and BRAF A694T pretreatment and had SD. At progression, cfDNA revealed an increase in NRAS VAF and mutations acquired in the MAPK pathway . One pt with prior FGFRi acquired FGFR2 V564I and V564K detected by cfDNA prior to initiation of TAS-120, and had a PR on TAS-120. There was a drop in FGFR2 V564I VAF with response that subsequently increased with progression. The patient also acquired a FGFR2 V564L mutation at progression. Conclusions: FGFR alterations can be detected by cfDNA. cfDNA may detect potential resistance mechanisms, including PI3K or MAPK pathway alterations and acquired FGFR2 mutations. Patients with gatekeeper mutations in cfDNA at baseline may still respond to TAS-120. Further study is needed to determine the impact of FGFR2 mutations and co-alterations on TAS-120 sensitivity.

A Model Information Management Plan for Molecular Pathology Sequence Data Using Standards: From Sequencer to Electronic Health Record

Incorporating genetic variant data into the electronic health record (EHR) in discrete computable fashion has vexed the informatics community for years. Genetic sequence test results are typically communicated by the molecular laboratory and stored in the EHR as textual documents. Although text documents are useful for human readability and initial use, they are not conducive for data retrieval and reuse. As a result, clinicians often struggle to find historical gene sequence results on a series of oncology patients within the EHR that might influence the care of the current patient. Second, identification of patients with specific mutation results in the EHR who are now eligible for new and/or changing therapy is not easily accomplished. Third, the molecular laboratory is challenged to monitor its sequencing processes for nonrandom process variation and other quality metrics. A novel approach to address each of these issues is presented and demonstrated. The authors use standard Health Level 7 laboratory result message formats in conjunction with international standards, Systematized Nomenclature of Medicine Clinical Terms and Human Genome Variant Society nomenclature, to represent, communicate, and store discrete gene sequence data within the EHR in a scalable fashion. This information management plan enables the support of the clinician at the point of care, enhances population management, and facilitates audits for maintaining laboratory quality.

Targeted multigene deep sequencing of Bruton tyrosine kinase inhibitor-resistant chronic lymphocytic leukemia with disease progression and Richter transformation

BACKGROUND

In a proportion of patients with chronic lymphocytic leukemia (CLL), resistance to Bruton tyrosine kinase (BTK) inhibitors (BTKi) is attributed to acquired BTK/phospholipase C gamma 2 (PLCG2) mutations. However, knowledge regarding additional genetic lesions associated with BTK/PLCG2 mutations, and gene mutations in patients lacking BTK/PLCG2 mutations, is limited.

METHODS

Using targeted deep sequencing, mutations in 29 genes associated with CLL and/or the BCR signaling pathway were assessed in patients with CLL who developed resistance to BTK inhibition with ibrutinib/acalabrutinib at a single institution.

RESULTS

The study group included 29 patients with BTKi-resistant CLL, 23 patients with disease progression, and 6 patients with Richter transformation (RT). The median times to disease progression and RT were 33.3 months and 13.3 months, respectively. In 11 patients, sequencing was possible at both baseline (prior to treatment with BTKi) and at time of disease progression/RT. Of these patients, 4 demonstrated BTK mutations at the time of disease progression/RT; patients without BTK mutations frequently acquired mutations associated with disease progression/RT in TP53, SF3B1, and CARD11, whereas additional mutations were rare in patients with BTK-mutated CLL. Sequencing of all 29 patients at the time of disease progression/RT identified BTK mutations at a frequency of 66%, including a novel V537I mutation. Among patients with disease progression, BTK mutations were observed in 16 patients (70%). The median time to disease progression was shorter in patients without BTK mutations compared with those with BTK-mutated CLL. Among patients with RT, SF3B1 mutations were more frequent than BTK mutations (67% vs 50%). Following BTKi discontinuation, we sequential mutation analysis was performed in 2 patients with RT and 3 patients with disease progression in the setting of persistent disease. Both patients with RT demonstrated disappearance of BTK and expansion of TP53 mutations. All 3 patients with disease progression received venetoclax and demonstrated suppression of BTK mutations.

CONCLUSIONS

Longitudinal, targeted, multigene deep sequencing is informative for the clinical monitoring of mutational evolution in patients with CLL who are receiving BTKi.

Hematologic malignancies and Li-Fraumeni syndrome

Li–Fraumeni syndrome (LFS) is an autosomal dominant condition associated with a high risk of a broad range of childhood- and adult-onset cancers. LFS is related to germline mutations of the tumor-suppressor gene TP53. The most common reported leukemia associated with LFS is hypodiploid acute lymphoblastic leukemia, but myeloid malignancies including acute myeloid leukemia (AML), chronic myeloid leukemia, and myelodysplastic syndrome (MDS) are also reported, often in the setting of therapy-related disease. We reviewed the clinicopathologic characteristics including cytogenetics and molecular analysis for seven adult patients with LFS and hematologic malignancies evaluated at the Hereditary Hematologic Malignancy Clinic (HHMC) at MD Anderson Cancer Center. We present this LFS review series to increase awareness of LFS for the appropriate diagnosis of both patients and potentially affected relatives, as well as provide experience with patient outcomes in this difficult to treat population.

P53 protein overexpression in de novo acute myeloid leukemia patients with normal diploid karyotype correlates with FLT3 internal tandem duplication and worse relapse-free survival

Although ~50% of acute myeloid leukemia (AML) patients have a normal diploid karyotype by conventional cytogenetics at diagnosis, this patient subset has a variable disease course and outcome. Aberrant overexpression of the p53 protein is usually associated with TP53 alterations and a complex karyotype, but the prevalence and impact of p53 overexpression in AML with diploid cytogenetics is unknown. We examined 100 newly diagnosed AML patients to evaluate the impact of p53 expression status quantified in bone marrow core biopsy samples using immunohistochemistry and computer-assisted image analysis. A total of 24 patients had p53 overexpression defined as 3+ staining intensity in ≥5% of cells; this finding correlated with lower platelet counts (P = .002), absence of CD34 expression in blasts (P = .009), higher bone marrow blast counts (P = .04), and a higher frequency of FLT3 internal tandem duplication (P = .007). Overexpression of p53 independently predicted for shorter leukemia-free survival in patients who underwent allogeneic stem cell transplantation by univariate (P = .021) and multivariate analyses (P = .004). There was no correlation between MDM2 and p53 protein expression in this cohort. We conclude that p53 expression evaluated by immunohistochemistry in bone marrow biopsy specimens at the time of AML diagnosis may indicate distinct clinical characteristics in patients with normal diploid cytogenetics and is a potentially valuable tool that can enhance risk-stratification.

A feasibility study of returning clinically actionable somatic genomic alterations identified in a research laboratory

PURPOSE

Molecular profiling performed in the research setting usually does not benefit the patients that donate their tissues. Through a prospective protocol, we sought to determine the feasibility and utility of performing broad genomic testing in the research laboratory for discovery, and the utility of giving treating physicians access to research data, with the option of validating actionable alterations in the CLIA environment.

EXPERIMENTAL DESIGN

1200 patients with advanced cancer underwent characterization of their tumors with high depth hybrid capture sequencing of 201 genes in the research setting. Tumors were also tested in the CLIA laboratory, with a standardized hotspot mutation analysis on an 11, 46 or 50 gene platform.

RESULTS

527 patients (44%) had at least one likely somatic mutation detected in an actionable gene using hotspot testing. With the 201 gene panel, 945 patients (79%) had at least one alteration in a potentially actionable gene that was undetected with the more limited CLIA panel testing. Sixty-four genomic alterations identified on the research panel were subsequently tested using an orthogonal CLIA assay. Of 16 mutations tested in the CLIA environment, 12 (75%) were confirmed. Twenty-five (52%) of 48 copy number alterations were confirmed. Nine (26.5%) of 34 patients with confirmed results received genotype-matched therapy. Seven of these patients were enrolled onto genotype-matched targeted therapy trials.

CONCLUSION

Expanded cancer gene sequencing identifies more actionable genomic alterations. The option of CLIA validating research results can provide alternative targets for personalized cancer therapy.

Fifteen-Year Journey to High Reliability in Pathology and Laboratory Medicine

Many high-reliability organizations in industries outside of health care have sustained high levels of excellence and prevention of harm while managing complex systems and risk. To date, no health care organizations has organized its efforts to achieve highly reliable results despite several decades of improvement science. Laboratorians were early adopters of quality initiatives and process improvements. In the late 1990s, the Division of Pathology and Laboratory Medicine at The University of Texas MD Anderson Cancer Center embarked on a major effort to improve quality and patient safety and to reduce waste. This article describes the institution's journey toward approaching high reliability with the intent to share not only the tools and best practices, but also the ongoing reassessment of the problems detected on the journey. The authors hope that their experience will help the reader develop interventions to adapt in their own environment to facilitate more optimal patient care.

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

640

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

Abstract A167: Targeting HER2 (ERBB2) amplification identified by next-generation sequencing (NGS) in patients with advanced or metastatic solid tumors

Introduction: Personalized cancer treatment is becoming more tumor agnostic by choosing a treatment based on the tumor genomics rather than the tumor type. HER2 is an effective therapeutic target with FDA-approved treatments in breast and gastric/gastroesophageal junction (GEJ) cancers; however, less is known about the efficacy of HER2-targeted treatment in other tumor types. Methods: Next-generation sequencing (NGS) was performed in 2221 patients (pts) with advanced solid tumors in CLIA-certified laboratories using multiple platforms for personalized cancer therapy. HER2 amplification (amp) was assessed by NGS platforms that report copy-number variations as per their respective algorithm. We assessed clinical characteristics and coalterations with HER2 amp. We evaluated the clinical benefit of HER2-targeted therapy, by measuring the progression-free survival (PFS) on HER2-matched targeted therapy (PFS2) compared to the PFS on prior therapy (PFS1). We also evaluated the response rate and overall survival (OS) of pts who received vs pts who did not receive HER2-targeted therapy. Results: A total of 122 pts (5.5%) were found to have HER2 amp. The most frequent tumor types were colorectal, biliary/gallbladder, gastric/gastroesophageal, esophageal, endometrial, head and neck squamous cell and salivary gland, non-small cell lung, and bladder cancers. Coalterations included mutations in TP53, APC, PIK3CA, LRP1B, NF1, KRAS, mutations and deletions in CDKN2A, and amp in MYC and CCNE1. Frequent amp in CDK12, RARA, and TOP2A amp (all within chromosome 17q) were also found in our pts. Concurrent mutations in HER2 were found in 16 pts (13%). Forty pts with HER2 amp on NGS also underwent HER2 IHC testing: 30 pts (75%) had overexpression (3+), 4 pts (10%) had equivocal expression (2+), 2 pts (5%) had low expression (1+), and 4 pts (10%) had no HER2 expression. FISH analysis was performed in 14 patients, out of which 12 patients were positive for amplification. Forty-four of 115 pts received at least 1 line of HER2-targeted therapy (range 1-4) with 42 pts receiving trastuzumab in combination with other drugs, including 11 (92%) of 12 pts with gastric, GEJ cancers with HER2 amp having received trastuzumab with chemotherapy (8 pts in the first line). Median OS of pts who received HER2-targeted therapy was 42 months vs 23 months for pts who did not receive HER2-targeted therapy (Hazard Ratio [HR] 0.6, 95% CI 0.38-0.97, p=0.0384). For 32 evaluable pts, PFS2/PFS1 ratio was ≥1.3 in 17 pts (53%) and ≥2 in 11 pts (34%) with median PFS2 of 23 weeks vs PFS1 of 11 weeks (p=0.0089). After exclusion of pts with gastric or GEJ cancers, pts receiving HER2-targeted therapy still had an improved OS (53 vs 23 months) (HR 0.56, 95% CI 0.33-0.93, p=0.0307) and the PFS2/PFS1 ratio was ≥1.3 in 15 (52%) of 29 pts with a median PFS2 of 23 weeks vs PFS1 of 12 weeks (p=0.0174). Conclusion: NGS reveals HER2 amp in a clinically relevant proportion of tumors and in a variety of tumor types. HER2-targeted therapy may confer clinical benefit in tumor types beyond those for which HER2 inhibitors are approved. The association of HER2 amp with genomic alterations in other oncogenic drivers provides rationale for novel therapeutic combinations.

Citation Format: Ecaterina Ileana Dumbrava, Kavitha Balaji, Kanwal Raghav, Milind Javle, Mariela Blum-Murphy, Blessy Sajan, Scott Kopetz, Russell Broaddus, Mark Routbort, Shubham Pant, Apostolia Tsimberidou, Vivek Subbiah, David S. Hong, Jordi Rodon Ahnert, Kenna Shaw, Sarina Piha-Paul, Funda Meric-Bernstam. Targeting HER2 (ERBB2) amplification identified by next-generation sequencing (NGS) in patients with advanced or metastatic solid tumors [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2017 Oct 26-30; Philadelphia, PA. Philadelphia (PA): AACR; Mol Cancer Ther 2018;17(1 Suppl):Abstract nr A167.

Clinical Use of Precision Oncology Decision Support

PURPOSE

Precision oncology is hindered by the lack of decision support for determining the functional and therapeutic significance of genomic alterations in tumors and relevant clinically available options. To bridge this knowledge gap, we established a Precision Oncology Decision Support (PODS) team that provides annotations at the alteration-level and subsequently determined if clinical decision-making was influenced.

METHODS

Genomic alterations were annotated to determine actionability based on a variant's known or potential functional and/or therapeutic significance. The medical records of a subset of patients annotated in 2015 were manually reviewed to assess trial enrollment. A web-based survey was implemented to capture the reasons why genotype-matched therapies were not pursued.

RESULTS

PODS processed 1,669 requests for annotation of 4,084 alterations (2,254 unique) across 49 tumor types for 1,197 patients. 2,444 annotations for 669 patients included an actionable variant call: 32.5% actionable, 9.4% potentially, 29.7% unknown, 28.4% non-actionable. 66% of patients had at least one actionable/potentially actionable alteration. 20.6% (110/535) patients annotated enrolled on a genotype-matched trial. Trial enrolment was significantly higher for patients with actionable/potentially actionable alterations (92/333, 27.6%) than those with unknown (16/136, 11.8%) and non-actionable (2/66, 3%) alterations (p=0.00004). Actionable alterations in PTEN, PIK3CA, and ERBB2 most frequently led to enrollment on genotype-matched trials. Clinicians cited a variety of reasons why patients with actionable alterations did not enroll on trials.

CONCLUSION

Over half of alterations annotated were of unknown significance or non-actionable. Physicians were more likely to enroll a patient on a genotype-matched trial when an annotation supported actionability. Future studies are needed to demonstrate the impact of decision support on trial enrollment and oncologic outcomes.

Abstract LB-103: Landscape of somatic ERBB2 Mutations: Findings from AACR GENIE and comparison to ongoing ERBB2 mutant basket study

Background: AACR GENIE is an international data-sharing project that aggregates clinical-grade cancer genomic data. As a demonstration of utility, we evaluated the landscape of ERBB2 mutations in the first 18,486 patients included in this registry and compared it to the first 100 patients enrolled in an ongoing international Phase 2 SUMMIT ‘basket’ study of the pan-HER inhibitor neratinib in ERBB2 mutant solid tumors (NCT01953926). Results: ERBB2 mutations were identified in 2.8% (519/18,486) of patients in the GENIE cohort and observed at all participating centers. In total, there were 482 missense, 66 indels, 19 truncating mutations, and 14 structural variants. A total of 263 unique missense mutations were observed including 12 at previously identified hotspots which accounted for 69.2% of all missense mutations. 35 unique cancer types were represented. The tumor types with the highest proportion of ERBB2 mutations were bladder (12.8%, 82/641), breast (3.9%, 87/2230), colorectal (3.3%, 70/2102), and NSCLC (3%, 90/3006). Among patients with copy number data available (91%) 11% had concurrent ERBB2 amplification, most often in breast cancer. The most frequently observed alterations in ERBB2, adjusted for differing exon coverage between panels, was S310F/Y in 0.46% of the GENIE cohort (12.6% of samples with ERBB2 alterations), Y772_A775dup in 0.21% (6.9%), R678Q in 0.17% (4.5%), L755S in 0.16% (5.2%), V777L in 0.12% (3.8%), and V842I in 0.09% (3.1%). The distribution of specific ERBB2 variants differed significantly by tumor type with exon 20 insertions being most common in NSCLC (44.4%, 40/90), L755S (18.9%, 11/92) in breast, S310F/Y (26.9%, 28/104) in bladder, and V842I (13.9%, 10/72) in colorectal cancer. Structural variants included intragenic deletions (n=4) and fusions involving various partners including GRB7 (n=2), and one each of C1orf87, PPIL6, HEXIM2, THRA, ASIC2, BCA3, WIPF2. The frequencies of ERBB2 mutant cancer types observed in the GENIE cohort were generally comparable to those enrolled to the neratinib basket study including NSCLC (17 vs 22%, respectively), breast (16.4 vs 24%), bladder (15.5 vs 14%), colorectal (13.2 vs 17%), and endometrial (4.2 vs 6%). At the variant level, S310F/Y was less prevalent in GENIE compared to the neratinib study (12.6 vs 24%) while all other mutations were generally similar including L755S (5.2 vs 9%), R678Q (4.5 vs 2%), Y772_A775dup (6.9 vs 13%), V777L (3.8 vs 9%), and V842I (3.1 vs 6%). Conclusion: GENIE confirms that a diversity of ERBB2 mutations are prevalent across a variety of tumor types in patients with advanced cancer. The genomic landscape of ERBB2 mutations was largely similar in the population based GENIE cohort and the neratinib SUMMIT study, providing the first direct evidence that basket study enrollment accurately reflects the true landscape of the target alteration.

Citation Format: Alison Schram, Helen H. Won, Fabrice Andre, Monica Arnedos, Funda Meric - Bernstam, Philippe L. Bedard, Kenna R. Shaw, Hugo Horlings, Christine Micheel, Ben Ho Park, Grace Mann, Alshad S. Lalani, Lillian Smyth, David B. Solit, Deborah Schrag, Mia A. Levy, Barrett J. Rollins, Mark Routbort, Charles L. Sawyers, Eva Lepisto, Michael F. Berger, David M. Hyman, on behalf of the AACR Project GENIE Consortium. Landscape of somatic ERBB2 Mutations: Findings from AACR GENIE and comparison to ongoing ERBB2 mutant basket study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr LB-103. doi:10.1158/1538-7445.AM2017-LB-103

Abstract LB-081: Novel and effective RUNX1-targeted therapy for AML expressing RUNX1 mutation

RUNX1 is a transcription factor involved in normal and malignant hematopoiesis. Somatic mutations in RUNX1 (mtRUNX1) have been documented in de novo and secondary AML (10%), MDS (~10%) and CMML (up to 37%). Germ-line, mono-allelic mutations and deletions in RUNX1 cause the highly-penetrant (~40%) autosomal dominant, Familial Platelet Disorder with a propensity to Myeloid Malignancy (FPD-MM). The majority of mtRUNX1 are missense mutations, large deletions or truncation-mutations in the DNA-binding ‘RUNT’ homology domain, or in the C-terminal transactivation domain. These mutations behave mostly as loss of function mutations, associated with relative resistance to standard chemotherapy and an unfavorable prognosis in AML. This highlights an unmet need to develop and test novel targeted therapies for AML due to germ-line or somatic mtRUNX1. In the present studies, we demonstrate that shRNA-mediated knockdown of mutant and wild-type RUNX1 repressed its target genes MYC, PU.1 and MPO (myeloperoxidase), as well as inhibited growth and induced apoptosis of AML cells expressing mtRUNX1 (OCI-AML5 and Mono-Mac-1). Ex vivo depletion of RUNX1 abrogated the leukemia initiating potential of OCI-AML5 cells. After engraftment, tetracycline-inducible shRNA-mediated in vivo knockdown of RUNX1, as compared to the non-induced controls, enhanced survival of the immune-depleted (NSG) mice engrafted with OCI-AML5 cells. RUNX1 transcription is driven by the BET protein BRD4-occupied super enhancer in the first intron of the RUNX1 gene. A heat map of Hi-C interaction scores within the RUNX1 TAD (topology-associated domain) showed triangle-shaped regions of high interaction, with CTCF binding sites defining the TAD boundaries, anchoring a loop that separates the RUNX1 TAD from other TADs. Consistent with this, our findings show that, shRNA-mediated depletion of BRD4 or treatment with BET protein (BETP) inhibitor (BETi) OTX015 reduced BRD4 occupancy at the enhancer and promoter of RUNX1. This was associated with depletion of RUNX1 and its target-gene expressions and apoptosis of cultured and patient-derived (PD), primary AML blast progenitor cells (BPCs). Additionally, treatment of NSG mice engrafted with luciferase-transduced OCI-AML5 cells with OTX015 (50 mg/kg/day X 5 days, for 3 weeks) reduced the AML burden and significantly improved their survival (p &lt; 0.01). Co-treatment with the BETi and BCL2-antagonist venetoclax or CDK4/6 antagonist palbociclib or decitabine synergistically induced apoptosis of OCI-AML5 and PD AML BPCs (combination index values of &lt; 1.0). Results describing the in vivo efficacy of BETi and BETP-PROTAC (proteolysis targeting chimera), alone and in combinations against AML expressing mtRUNX1 will be updated at the AACR meetings. Collectively, these findings highlight novel and effective, mechanistically-targeted, single agent or combination, BETP antagonist-based therapy of AML expressing mtRUNX1.

Citation Format: Christopher P. Mill, Courtney DiNardo, Warren C. Fiskus, Dyana T. Saenz, Baohua Sun, Agnieszka J. Nowak, Misun Kim, Mark Routbort, Koichi Takahashi, Kapil N. Bhalla. Novel and effective RUNX1-targeted therapy for AML expressing RUNX1 mutation [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr LB-081. doi:10.1158/1538-7445.AM2017-LB-081

Prevalence of incidental germline pathogenic (PV) and likely pathogenic (LPV) variants in hereditary cancer-related genes identified in matched tumor/normal sequencing of advanced solid tumors

1524

Background: Next-generation sequencing (NGS) for tumor molecular profiling can reveal germline incidental mutations in hereditary cancer-related genes. The American College of Medical Genetics and Genomics (ACMG) has recommended that laboratories performing clinical sequencing seek and report PV and LPV in 56 genes. We assessed the prevalence of incidental germline LPV and PV in other cancer-related genes among patients undergoing hybrid capture sequencing of 201 cancer-related genes. Methods: Matched tumor and germline DNA NGS of a targeted panel of 201 genes was performed in 1000 patients (pts) with advanced or metastatic solid tumors enrolled in a molecular testing protocol (NCT01772771) in a research laboratory. We previously reported germline alterations in the putative most actionable genes as designated by ACMG (PMID: 26787237). We assessed the germline LPV and PV in 54 additional cancer-related genes. Results: Among the 1000 patients who underwent tumor and normal DNA sequencing, 37 patients (3.7%) were found to have a germline PV or LPV in the following genes: ATM (4); BAP1 (1); CDH1 (1); CDKN2A (1); CHEK1 (2); CHEK2 (10); EGFR (1); ERCC3 (4); ERCC5 (1); HNF1B (1); HRAS (1); MLL3 (1); NF1 (3); PKHD1 (4); PTCH1 (1) and SMARCA4 (1). Eight pts (22%) had previous genetic counseling and testing for various reasons, but only 3 pts (8%) had previously identified alterations (all with NF1 mutations). After discussion in our return of germline results board, it was decided to return the findings in established hereditary cancer predisposition genes with high penetrance: BAP1 (p.Y401X), CDH1 (p.C688X), CDKN2A (p.G101W), EGFR (p.T790M) and SMARCA4 (p.S332FfsX55) after validation in a CLIA laboratory. Conclusions: Return of the previously unrecognized germline LPV or PV in patients with advanced or metastatic cancers who undergo somatic profiling is of great interest. The exact genes for which the germline results should be returned is controversial. Broader genomic testing is likely to identify additional incidental germline alterations with potential clinical utility to patients and their relatives.

Molecular characterization of TP53 mutations and copy number change in colorectal cancers

e15143

Background: TP53 mutations (mut) are prevalent in colorectal cancer (CRC) patients (pts) and can result in oncogenic gain of function effect as well as loss of tumor suppressor function depending on the mut. Methods: The Cancer Genome Atlas (TCGA) data of 220 CRC pts with tumors sequenced and gene copy numbers (CN) annotated was obtained. Pts with CMS subtype data also available were included (n = 167). Chi square test was used to compare frequency of TP53 mut and CN change among CMS subtypes. TP53 mRNA and protein levels were compared by TP53 mut and CN change using independent sample t test. Results: 86 (51.5%) pts were TP53 mutant, with 35.3% missense and 16.2% loss of function (LOF) mut. LOF mut included 14 non-sense and 13 frameshift mut, and occurred mostly in codons 213 and 306. Missense mut mostly affected codons 175, 248 and 273 in 10, 9 and 8 pts respectively. TP53 mut were most frequent in CMS2 and CMS4 subtypes (62.7 and 62.5% respectively). TP53 mut in CMS2 were mostly missense and CMS 4 pts mostly had LOF mut. TP53 CN loss was seen in 74.7% CMS 2, 68.8% CMS 4, 43.5% CMS 3 and 6.9% CMS1 pts. 84.9% pts with TP53 mut had CN loss vs 28.4% of wild type pts (p < 0.001). Frequency of TP53 CN loss did not vary significantly by type of TP53 mut. Pts with TP53 missense mut had higher p53 mRNA and protein levels than those with LOF mut. Pts with TP53 CN loss had lower mRNA levels but no change in protein levels than those with normal CN. Conclusions: The higher TP53 mRNA and protein levels with missense mut suggest possible gain of function. Research on interactions of TP53 change with wnt, myc, and TGF beta pathway genes may reveal synthetic lethal treatment combinations. [Table: see text]

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

3589

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

RAS heterogeneity as a prognostic marker in metastatic colorectal cancer

586

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

Clinical next generation sequencing to identify actionable aberrations in a phase I program

Purpose

We determined the frequency of recurrent hotspot mutations in 46 cancer-related genes across tumor histologies in patients with advanced cancer.

Methods

We reviewed data from 500 consecutive patients who underwent genomic profiling on an IRB-approved prospective clinical protocol in the Phase I program at the MD Anderson Cancer Center. Archival tumor DNA was tested for 740 hotspot mutations in 46 genes (Ampli-Seq Cancer Panel; Life Technologies, CA).

Results

Of the 500 patients, 362 had at least one reported mutation/variant. The most common likely somatic mutations were within TP53 (36%), KRAS (11%), and PIK3CA (9%) genes. Sarcoma (20%) and kidney (30%) had the lowest proportion of likely somatic mutations detected, while pancreas (100%), colorectal (89%), melanoma (86%), and endometrial (75%) had the highest. There was high concordance in 62 patients with paired primary tumors and metastases analyzed. 151 (30%) patients had alterations in potentially actionable genes. 37 tumor types were enrolled; both rare actionable mutations in common tumor types and actionable mutations in rare tumor types were identified.

Conclusion

Multiplex testing in the CLIA environment facilitates genomic characterization across multiple tumor lineages and identification of novel opportunities for genotype-driven trials.