Virtual Library

Start Your Search

Fred R Hirsch



Author of

  • +

    MA08 - Advances in Biomarkers for Immune Checkpoint Blockade and Targeted Therapy in Non Small Cell Lung Carcinoma (ID 166)

    • Event: WCLC 2020
    • Type: Mini Oral
    • Track: Pathology, Molecular Pathology and Diagnostic Biomarkers
    • Presentations: 1
    • +

      MA08.10 - LUNGMAP Master Protocol (LUNGMAP): Concordance Between Plasma ctDNA and Tissue Molecular Analysis (ID 3146)

      16:45 - 17:45  |  Author(s): Fred R Hirsch

      • Abstract
      • Slides

      Introduction

      The national LUNGMAP clinical trial is predicated on molecular screening enabling patient enrollment to biomarker-matched sub-studies for rapid evaluation of new precision medicine concepts in advanced NSCLC. To date, LUNGMAP has used a tissue-based Next-Generation Sequencing (NGS) approach for biomarker assessment. Given the utility of circulating tumor DNA (ctDNA) for biomarker identification, LUNGMAP investigators are evaluating the feasibility of plasma ctDNA as a screening approach.

      Methods

      Plasma samples for ctDNA testing were required for patients submitting fresh tissue biopsies for LUNGMAP screening. Tissue and plasma ctDNA were analyzed using the FoundationONE CDx and FoundationACT platforms at Foundation Medicine, Inc., respectively. Alterations detectable in both platforms were evaluated. Using tissue-detected driver alterations (referred to as drivers) as the gold standard, sensitivity was calculated as the proportion of patients with drivers also detected in ctDNA in addition to tissue, and specificity was calculated as the proportion of patients without drivers in ctDNA among those without drivers in tissue. Proportions and 95% exact confidence interval (CI) estimates were calculated.

      Results

      From January 2019 to June 2020, 129 patients had paired data and 54 (42%) had recognized oncogene drivers detected (EGFR [n=7], KRAS [n=37], MET [n=7], RET [n=2], BRAF [n=1], Table 1). Fifty-two patients had drivers detected in tissue; of these 43 were also observed in ctDNA, with 9 found in tissue only, for a ctDNA driver sensitivity of 83% (43/52, 95% CI: 74-93%). Of the 77 patients with no drivers in tissue, 2 drivers were detected in ctDNA (EGFR Ex20ins, MET amp) for a ctDNA specificity of 97% (75/77, 95% CI: 91-100%). For drivers, median variant allele frequency (VAF) in ctDNA was 2.22% (range: 0.13%-46.27%). For all single nucleotide variants (SNVs) and rearrangements detectable on both platforms, 386 variants were detected. Short variants (point mutations and small in/dels) showed the most fidelity, with 54% detected in both platforms (Table 1). Copy number alterations using an earlier platform version were least reproduced, with 8% identified by both.

      Conclusion

      In the LUNGMAP population, ctDNA (FoundationAct) had an 83% sensitivity and 97% specificity for NSCLC drivers detected in tissue. For non-driver alterations, additional variants were detected exclusively in plasma or tissue, likely reflecting differential sensitivity and/or non-shedding and tissue heterogeneity. These results, consistant with other recent studies, support the planned use of ctDNA for enrollment onto LUNGMAP sub-studies, with a positive finding meriting inclusion in study but a negative finding, considered inconclusive, requiring use of tissue results.

      Table 1

      N (%)

      Total Alterations Detected

      Number of Patients

      ................... In ctDNA ................

      ...................... In Tissue ................

      Overall

      In Tissue

      Not in Tissue

      Overall

      In ctDNA

      Not in ctDNA

      Driver Alterations

      54

      54

      45

      43 (96%)

      2 (4%)

      52

      43 (83%)

      9 (17%)

      Non-driver Alterations

      439

      75

      294

      169 (57%)

      125 (43%)

      314

      169 (54%)

      145 (46%)

      Short Variants

      316

      273

      158 (58%)

      115 (42%)

      201

      158 (79%)

      43 (21%)

      Copy Number Alts

      104

      10

      8 (80%)

      2 (20%)

      102

      8 (8%)

      94 (92%)

      Rearrangements

      19

      11

      3 (27%)

      8 (73%)

      11

      3 (27%)

      8 (73%)

      Overall

      493

      129

      339

      212 (63%)

      127 (37%)

      366

      212 (58%)

      154 (42%)

      Short Variants

      365

      314

      198 (63%)

      116 (37%)

      249

      198 (80%)

      51 (20%)

      Copy Number Alts

      107

      12

      9 (75%)

      3 (25%)

      104

      9 (9%)

      95 (91%)

      Rearrangements

      21

      13

      5 (38%)

      8 (62%)

      13

      5 (38%)

      8 (62%)

      TP53

      150

      128

      77 (60%)

      51 (40%)

      99

      77 (78%)

      22 (22%)

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

  • +

    MA11 - Expanding Targetable Genetic Alterations in NSCLC (ID 251)

    • Event: WCLC 2020
    • Type: Mini Oral
    • Track: Targeted Therapy - Clinically Focused
    • Presentations: 1
    • +

      MA11.10 - Lung Master Protocol (Lung-MAP) Next Generation Sequencing Analysis of Advanced Squamous Cell Cancers (SWOG S1400) (ID 3055)

      14:15 - 15:15  |  Author(s): Fred R Hirsch

      • Abstract
      • Slides

      Introduction

      SWOG S1400, the original screening protocol of Lung-MAP, enrolled patients with Stage IV or recurrent squamous cell lung cancer previously treated with at least one line of systemic therapy. Tumors were profiled by NGS using Foundation Medicine’s FoundationOne T5 research platform, which sequenced the exons and/or introns of 313 cancer-related genes. Here, we report the results of a comprehensive analysis of the S1400 NGS data compared to The Cancer Genome Atlas (TCGA) data, including identification of novel sets of mutually exclusive and co-occurring genetic alterations.

      Methods

      Analyses included all patients with successful NGS testing enrolled on S1400. Mutually Exclusive Gene Set Analysis (MEGSA) was used to identify sets across genetic alterations with mutated prevalence > 6%. Selected Events Linked by Evolutionary Conditions across human Tumors (SELECT) was used to identify pairwise gene interactions. Comparisons were performed using mutation profiles of 495 squamous cell lung cancers downloaded from the TCGA data portal. Cox proportional hazards models adjusted for clinical covariates including age, sex, smoking history and AJCC TNM categories were used to examine the association between each genetic variant and survival. The Benjamini-Hochberg method was used to adjust significance values for multiple comparisons.

      Results

      Between June 16, 2014 and January 29, 2019, 1864 patients were enrolled to be screened, of whom NGS was available for 1672. 73% of the sequenced tumor samples were archival and 27% were fresh biopsies; there were no significant differences in prevalence of genetic alterations between these. MEGSA identified two non-overlapping sets of mutually exclusive gene alterations with a false discovery rate (FDR) < 15%: NFE2L2, KEAP1 and PARP4 (FDR = 4.1%) and CDKN2A and RB1 (FDR = 13.1%). Mutual exclusivity of NFE2L2 and KEAP1 alterations has been previously observed, e.g., in TCGA, however mutual exclusivity of PARP4 and NFE2L2 or KEAP1 alterations is a novel finding. SELECT identified 41 pairs of mutually exclusive and 95 pairs of co-occurring gene alterations. Top significant co-occurring pairs that appeared in this dataset but not TCGA include CDKN2A and TP53, KRAS and STK11, HGF and MLL2, PDGFRB and SMARCA4, NFE2L2 and TP53, ATRX and RUNX1T1, GRIN2A and NCOR1, and MCL1 and MYCN. Male sex and smoking history were associated with poorer survival. When these and other clinical covariates were incorporated in Cox proportional hazards models, there were no individual genetic variants that were associated with survival; however, NFE2L2 and KEAP1 alterations when taken together were associated with poorer survival.

      Conclusion

      This analysis of the Lung-MAP S1400 NGS data features a substantially larger sample size than any previously published dataset of squamous cell lung cancers, although it is limited to genes sequenced on the FoundationOne T5 platform. Compared to TCGA, this dataset features a homogeneous set of subjects all with previously treated advanced disease and enrolled on a clinical trial. Novel findings, including mutual exclusivity of PARP4 and NFE2L2 or KEAP1 alterations, suggest that PARP4 may have a hitherto undiscovered role in a key pathway known to impact responses to oxidative stress and treatment resistance.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

  • +

    OA01 - Established Drugs in Special Populations and New Drugs in Established Populations (ID 226)

    • Event: WCLC 2020
    • Type: Oral
    • Track: Immunotherapy (Phase II/III Trials)
    • Presentations: 1
    • +

      OA01.04 - Tumor Mutation Burden (TMB) by Next Generation Sequencing (NGS) Associates with Survival (OS) in Lung-MAP Immunotherapy Trials S1400I and S1400A (ID 3229)

      09:15 - 10:15  |  Presenting Author(s): Fred R Hirsch

      • Abstract
      • Presentation
      • Slides

      Introduction

      TMB is an emerging biomarker for efficacy of immune checkpoint inhibitors (ICI). Lung-MAP is a master protocol for biomarker-driven trials in advanced NSCLC. Two sub-studies in previously treated ICI naïve advanced squamous cell lung cancer (sqNSCLC), S1400I, a phase III trial randomizing patients to nivolumab plus ipilimumab (N/I) versus nivolumab (N), and S1400A, a phase II trial evaluating durvalumab (D), provided the opportunity to evaluate TMB and related biomarkers by NGS and to determine associations with clinical outcomes.

      Methods

      NGS was performed on DNA from formalin-fixed paraffin-embedded tumor specimens using the FoundationOne T5 platform. TMB was defined as the total number of nonsynonymous mutations per megabase (Mb) of coding sequence. In S1400I, PD-L1 expression was assessed by the 22C3 antibody. A Cox model was used to evaluate associations between TMB (continuous and dichotomized at 10 Mb/mt), PD-L1 (continuous and dichotomized at 0% versus > 0%), overall survival (OS) and progression-free survival (PFS), summarized by hazard ratios (HRs) and 95% confidence intervals (CI). Associations between TMB and genetic alterations were evaluated by Wald test, with false discovery rate (FDR) ≤ 5% scored as positive. Unsupervised hierarchical clustering was performed using combined data from S1400I+S1400A.

      Results

      3229 figure.jpg252 patients on N/I or N and 68 patients on D were included in the analysis. Higher TMB (per 10-unit difference in TMB value) was significantly associated with better OS and PFS (OS HR(CI): 0.80 (0.67–0.94), P = 0.008 and PFS HR(CI): 0.80 (0.69–0.93), P = 0.004). In S1400I, PD-L1 expression levels were not significantly associated with OS or PFS (N=161, P > 0.05), alone or in combination with TMB. In S1400I+S1400A, no genetic variants were significantly associated with OS or PFS. Genes whose alterations were significantly associated with TMB are shown in the volcano plot. Unsupervised hierarchical clustering suggested a variant-defined subgroup conferred better PFS (HR (CI): 0.41 (0.19–0.88), P = 0.022) but not OS; notably, this subgroup showed 3.8-fold higher TMB and more frequent alterations of genes shown in the plot, compared to other subgroups.

      Conclusion

      Several different methodologies have been employed to measure TMB. TMB by FoundationOne NGS is an analytically and clinically validated assay correlating well with WES and predicted neoantigen load. Here we report that high TMB, but not PD-L1, is associated with improved OS and PFS in patients treated with ICI on S1400I/S1400A. How genetic alterations associated with high TMB may biologically contribute to clinical outcomes from ICI warrants further consideration.

      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.

  • +

    P35 - Pathology - Genomics (ID 105)

    • Event: WCLC 2020
    • Type: Posters
    • Track: Pathology, Molecular Pathology and Diagnostic Biomarkers
    • Presentations: 1
    • Moderators:
    • Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
    • +

      P35.09 - Oncogenetic Differences in Never-Smokers versus Smokers with NSCLC Adenocarcinoma Treated at the Mt Sinai Tisch Cancer Institute (ID 3423)

      00:00 - 00:00  |  Author(s): Fred R Hirsch

      • Abstract
      • Slides

      Introduction

      Never-smokers with lung adenocarcinoma (LUAD) have high rates of actionable driver mutations; however, their scope, distribution and clinical utility requires more quantification.

      Methods

      All Mount Sinai patients between 01/01/2015 and 6/01/2020 with a diagnosis of LUAD of any stage with known smoking status who received genetic testing were included. Most patients were analyzed using the Ion or Oncomine hotspot panels or the OCAv3 NGS panel conducted at Sema4. A driver abnormality is defined here as an activating mutation in a signal transduction factor on which the tumor is singularly dependent on for growth and or survival. Drivers were categorized as Tier1: matched with FDA-approved treatment, Tier2: matched with investigational or off-label treatment, and Tier3: non-targetable with current treatments. Patients were considered fully genotyped if they were comprehensively analyzed for alterations in EGFR, KRAS, MET, ALK, RET, ROS1, BRAF, NTRK1-3, ERBB2, NRAS and HRAS, or were considered partially genotyped otherwise.

      Results

      258 never-smokers and 705 smokers were identified as meeting the above criteria. Of the never-smokers, 214 (83%) had a driver mutation with 189 (73%) considered actionable (Tier1 or 2). Within the smoker cohort, 460 (65%) had an identified driver mutation with 305 (43%) actionable (p<0.0001) (Figure 1). When fully and comprehensively sequenced, 95% (73/77) of never-smokers had a driver mutation with 81% (62/77) actionable; whereas, for smokers, 73% (144/197) had a driver with only 50% (99/197) actionable (p<0.0001). Significant differences in the rates and subtypes of EGFR, ALK and KRAS alterations were observed between never-smokers and smokers (Table 1).

      Table 1. Significant differences by smoking (P-values: Fisher's Exact Test)

      Never Smokers (N = 258)

      Smokers (N = 705)

      P-value

      Total EGFR Drivers

      139 (54%)

      120 (17%)

      <0.0001

      Total KRAS Drivers

      32 (12%)

      264 (37%)

      <0.0001

      ALK Fusions

      16 (6.2%)

      11 (1.5%)

      < 0.001

      Any Oncogenic Driver

      214 (83%)

      460 (65%)

      <0.0001

      Any Actionable Driver

      189 (73%)

      305 (43%)

      <0.0001

      Any FDA-indicated Driver

      164 (63%)

      157 (22%)

      <0.0001

      KRAS G12C / All KRAS

      6 / 32 (18%)

      112 / 264 (42%)

      0.012

      EGFR Ex19del / All EGFR

      71 / 139 (51%)

      46 / 120 (38%)

      0.046

      EGFR G719x / All EGFR

      5 / 139 (3.6%)

      19 / 120 (16%)

      < 0.001

      abstract_fig_with_colors.png

      Conclusion

      Actionable driver mutations were frequently observed in both smokers and never-smokers, but were significantly more numerous in never-smokers. Comprehensive NGS revealed driver alterations in 95% of never-smokers. In summary, all efforts should be exhausted to identify or rule out the presence of an actionable driver mutation in LUAD patients requiring systemic treatment.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.