- Email: [email protected]
Comprehensive genomic profiling and outcomes among metastatic melanoma patients (pts) treated with first-line cancer immunotherapy (CIT) in a real-world setting
abstracts 4O
Tumour cells mirror embryonic developmental programs to acquire invasive and metastatic capabilities
S. Saghafinia, I. Michael, K. Homicsko, D. Hanahan EPFL - Swiss Institute for Experimental Cancer Research (ISREC), Lausanne, Switzerland
5O
Using single cell data to validate the cellular origins of a clonal expression biomarker in lung cancer
D. Biswas1, N.J. Birkbak2, N. McGranahan3, C. Swanton4 UCL Cancer Institute/Paul O’Gorman Building, London, UK, 2Aarhus University, Aarhus, Denmark, 3Cancer Research UK, London, UK, 4Translational Cancer Therapeutics, The Francis Crick Institute, London, UK
1
Background: Molecular biomarkers aim to stratify cancer patients into disease subtypes predictive of outcome, improving diagnostic precision beyond clinical descriptors such as tumour stage. Transcriptomic intra-tumour heterogeneity (RNA-ITH) has been shown to confound existing expression-based biomarkers across multiple cancer types. Methods: Here, we analyse multi-region whole-exome and RNA sequencing data for 156 tumour regions from 48 TRACERx patients to explore and control for RNA-ITH in non-small cell lung cancer (NSCLC). Results: We find that chromosomal instability (CIN) is a major driver of RNA-ITH, and existing prognostic gene expression signatures are vulnerable to tumour sampling bias. To address this, we identify genes expressed homogeneously within individual tumours that encode expression modules of cancer cell proliferation and are often driven by DNA copy-number gains selected early in tumour evolution. Leveraging single-cell data, we examine the origins of the expression signals, and relate our findings to published biological and clinical gene expression signatures. Conclusions: Clonal transcriptomic biomarkers overcome tumour sampling bias, associate with survival independently of clinicopathological risk factors, and may provide a general strategy to refine biomarker design across cancer types. Clinical trial identification: NCT0188860. Legal entity responsible for the study: University College London (UCL/12/0279) and has been approved by an independent Research Ethics Committee (13/LO/1546). Funding: Francis Crick Institute that receives its core funding from Cancer Research UK (FC001169, FC001202), the UK Medical Research Council (FC001169, FC001202), and the Wellcome Trust (FC001169, FC001202). The research leading to these results has received funding from the European Research Council (ERC) under the European Union’s Seventh Framework Programme (FP7/2007-2013) Consolidator Grant (FP7-THESEUS-617844), European Commission ITN (FP7-PloidyNet 607722), ERC Advanced Grant (PROTEUS) has received funding from the European Research
vii2 | Molecular Analysis for Personalised Therapy
Council under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 835297), Chromavision – this project has received funding from the European’s Union Horizon 2020 research and innovation programme (grant agreement No. 665233). Disclosure: D. Biswas: Non-remunerated activity/ies, patent co-inventor: CRUK; Studentship: Jean Shanks Foundation MBPhD; Funding (self): MBPhD programme at University College London; Funding (institution): the NIHR BRC at University College London Hospitals. N.J. Birkbak: Nonremunerated activity/ies, patent co-inventor: CRUK; Fellow: Lundbeck Foundation; Funding (self): the Aarhus University Research Foundation; Funding (institution): Danish Cancer Society. N. McGranahan: Non-remunerated activity/ies, patent co-inventor: CRUK. Fellow: Sir Henry Dale Fellow, jointly funded by the Wellcome Trust and the Royal Society (Grant Number 211179/Z/18/Z); Funding (institution): CRUK, Rosetrees, the NIHR BRC at University College London Hospitals. C. Swanton: Advisory / Consultancy, Research grant / Funding (institution): Pfizer; Advisory / Consultancy, Research grant / Funding (institution): AstraZeneca; Advisory / Consultancy, Research grant / Funding (institution): BMS; Advisory / Consultancy, Research grant / Funding (institution): Roche; Advisory / Consultancy, Research grant / Funding (institution): Ventana; Advisory / Consultancy: Novartis; Advisory / Consultancy: GlaxoSmithKline; Advisory / Consultancy: MSD; Advisory / Consultancy: Celgene; Advisory / Consultancy: Illumina; Advisory / Consultancy: Genentech; Advisory / Consultancy, Shareholder / Stockholder / Stock options: GRAIL; Advisory / Consultancy: Medicxi; Advisory / Consultancy: Sarah Cannon Research Institute; Advisory / Consultancy: Dynamo Therapeutics; Shareholder / Stockholder / Stock options: ApoGen Biotechnologies; Shareholder / Stockholder / Stock options: Epic Bioscience; Shareholder / Stockholder / Stock options: Achilles Therapeutics. Funding (institution) Cancer Research UK (TRACERx and CRUK Cancer Immunotherapy Catalyst Network), the CRUK Lung Cancer Centre of Excellence, Stand Up 2 Cancer (SU2C), the Rosetrees Trust, Butterfield and Stoneygate Trusts, NovoNordisk Foundation (ID16584), the Prostate Cancer Foundation, the Breast Cancer Research Foundation (BCRF), National Institute for Health Research, the University College London Hospitals Biomedical Research Centre, and the Cancer Research UK University College London Experimental Cancer Medicine Centre.
6O
Comprehensive genomic profiling and outcomes among metastatic melanoma patients (pts) treated with first-line cancer immunotherapy (CIT) in a real-world setting
N. Sadetsky1, P. Lambert2, C. Julian2, J. Chen2, Y. Yan1 Genentech Inc., Roche USA, South San Francisco, CA, USA, 2Genentech Inc., South San Francisco, CA, USA
1
Background: Distinctive genomic subtypes (BRAF-mutated [BRAFmut], NRASmut, NF1mut, and triple wild type [wt]) have been identified in melanoma, but little is known about their distribution and association with outcomes outside of clinical trials. By linking longitudinal electronic health records (EHR) and comprehensive genomic profiling we aim to describe characteristics and outcomes among CIT-treated pts in a real-world setting. Methods: Metastatic melanoma pts in the de-identified Flatiron Health (FH)Foundation Medicine (FMI) Clinico-Genomic Database (CGDB), in which EHRderived data from FH are linked to genomic data from FMI, who received first-line pembrolizumab (pembro), nivolumab (nivo), ipilimumab (ipi), or ipi þ nivo between Jan 1, 2011, and Nov 30, 2018, were analyzed. Median OS was calculated from the start of first-line therapy and was evaluated according to BRAF status and genomic subtypes. Results: Of 656 melanoma pts in CGDB, 236 received first-line CIT. The majority of pts (69%, n ¼ 165) were BRAFwt and the rest (31%, n ¼ 71) were BRAFmut. Median age was 62.9 yrs (57.3 yrs for BRAFmut pts and 65.6 yrs for BRAFwt pts), 89% were white, and 97% were treated in community clinics. Among BRAFwt pts, 40% (n ¼ 66) were NRASmut, 32% (n ¼ 53) were NF1mut, and 28% (n ¼ 46) were triple wt. Pembro was used in 36% of pts, followed by ipi þ nivo (25%), nivo (20%), and ipi (18%). Almost half (47%) of the pts received subsequent therapy; of these pts, 53% (n ¼ 59) received CIT, 23% (n ¼ 26) received targeted therapy, and 24% (n ¼ 27) received other therapies. Median OS for BRAFwt and BRAFmut pts after first-line CIT was 38.6 mo (95% CI 20.2–not estimable [NE]) and 28.9 mo (95% CI 23.3–NE), respectively. Among BRAFwt pts, median OS was 44.9 mo (95% CI 28.9–NE) for NRASmut pts, 27.1 mo (95% CI 19.4–NE) for NF1mut pts, and 19.8 mo (95% CI 11.8-NE) for triple wt pts. Conclusions: Our study demonstrated that outcomes among melanoma pts receiving first-line CIT in real-world setting vary based on genomic subtypes and revealed, for the first time, differences in outcomes of major genomic subtypes in BRAFwt pts. Continued investigation of the association between specific genomic subtypes and survival in a real-world setting is needed. Editorial acknowledgement: Melanie Sweetlove, MSc (ApotheCom, Yardley, PA, USA). Legal entity responsible for the study: Genentech, a member of the Roche Group. Funding: Genentech, a member of the Roche Group. Disclosure: N. Sadetsky: Shareholder / Stockholder / Stock options, Full / Part-time employment: Genentech, a member of the Roche Group. P. Lambert: Shareholder / Stockholder / Stock options, Full / Part-time employment: Genentech, a member of the Roche Group. C. Julian: Shareholder / Stockholder / Stock options, Full / Part-time employment: Genentech, a member of the Roche Group. J. Chen: Shareholder / Stockholder / Stock options, Full / Part-time employment: Genentech, a member of the Roche Group. Y. Yan: Shareholder / Stockholder / Stock options, Full / Part-time employment: Genentech, a member of the Roche Group.
Volume 30 | Supplement 7 | November 2019
Downloaded from https://academic.oup.com/annonc/article-abstract/30/Supplement_7/mdz413.011/5613278 by guest on 24 December 2019
Background: Pancreatic Neuroendocrine Tumours (PanNETs) are the second most common form of pancreatic cancer. We previously identified two principal subtypes of PanNET: insulinomas (islet tumors; IT) and metastasis-like primaries (MLP), corresponding to the low- and high-grade classification of human PanNETs. This study describes the mechanisms by which PanNETs progress from IT to more aggressive MLP tumors and eventually metastasize. Methods: We profiled single cell transcriptomes, bulk mRNA and miRNA transcriptomes and the proteomes of primary and metastasis specimens from the genetically engineered mouse model of PanNETs (Rip1 Tag2). Results: We found that the IT tumours maintained the expression of mature ß cell markers. In contrast, the MLP tumours expressed pancreatic progenitor markers. By integrating the data on ß-cell differentiation from pancreatic progenitors to mature ß cells, we demonstrated that the tumour progression from IT to MLP follows the reverse embryonic and postnatal developmental path. Furthermore, we identified Hmgb3 and miR-181cd cluster as the MLP master regulators. Over-expression of the miR-181cd cluster in IT-like cancer cell-lines resulted in the acquisition of the MLP gene signature and MLP morphologic phenotypes, in addition to activation Hmgb3 expression. This suggested a central role for Hmgb3 in initiating the MLP program. Furthermore, inhibiting the expression of Hmgb3 in MLP-like cancer cell-lines resulted in a significant growth decrease, demonstrating the importance of Hmgb3 in the maintenance of MLP-like cell state in vitro. Using transcriptomic data from human patients, we evaluated the relevance of the MLP program in human tumours. We established that aggressive human PanNET tumours also follow the same reverse developmental trajectory of dedifferentiation. Notably, patients with high MLP genes expression had a worse prognosis. Conclusions: These data demonstrate dedifferentiation as a mechanism by which malignant neuroendocrine cancer cells acquire progenitor-like features, enabling them to become more aggressive and metastatic. In addition, miR-181cd cluster and Hmgb3 act as the core regulators in the initiation of dedifferentiation and maintenance of progenitor-like features in tumour cells. Legal entity responsible for the study: CMSO Group at ISREC, EPFL. Funding: Swiss National Science Foundation. Disclosure: All authors have declared no conflicts of interest.
Annals of Oncology
Tumour cells mirror embryonic developmental programs to acquire invasive and metastatic capabilities
S. Saghafinia, I. Michael, K. Homicsko, D. Hanahan EPFL - Swiss Institute for Experimental Cancer Research (ISREC), Lausanne, Switzerland
5O
Using single cell data to validate the cellular origins of a clonal expression biomarker in lung cancer
D. Biswas1, N.J. Birkbak2, N. McGranahan3, C. Swanton4 UCL Cancer Institute/Paul O’Gorman Building, London, UK, 2Aarhus University, Aarhus, Denmark, 3Cancer Research UK, London, UK, 4Translational Cancer Therapeutics, The Francis Crick Institute, London, UK
1
Background: Molecular biomarkers aim to stratify cancer patients into disease subtypes predictive of outcome, improving diagnostic precision beyond clinical descriptors such as tumour stage. Transcriptomic intra-tumour heterogeneity (RNA-ITH) has been shown to confound existing expression-based biomarkers across multiple cancer types. Methods: Here, we analyse multi-region whole-exome and RNA sequencing data for 156 tumour regions from 48 TRACERx patients to explore and control for RNA-ITH in non-small cell lung cancer (NSCLC). Results: We find that chromosomal instability (CIN) is a major driver of RNA-ITH, and existing prognostic gene expression signatures are vulnerable to tumour sampling bias. To address this, we identify genes expressed homogeneously within individual tumours that encode expression modules of cancer cell proliferation and are often driven by DNA copy-number gains selected early in tumour evolution. Leveraging single-cell data, we examine the origins of the expression signals, and relate our findings to published biological and clinical gene expression signatures. Conclusions: Clonal transcriptomic biomarkers overcome tumour sampling bias, associate with survival independently of clinicopathological risk factors, and may provide a general strategy to refine biomarker design across cancer types. Clinical trial identification: NCT0188860. Legal entity responsible for the study: University College London (UCL/12/0279) and has been approved by an independent Research Ethics Committee (13/LO/1546). Funding: Francis Crick Institute that receives its core funding from Cancer Research UK (FC001169, FC001202), the UK Medical Research Council (FC001169, FC001202), and the Wellcome Trust (FC001169, FC001202). The research leading to these results has received funding from the European Research Council (ERC) under the European Union’s Seventh Framework Programme (FP7/2007-2013) Consolidator Grant (FP7-THESEUS-617844), European Commission ITN (FP7-PloidyNet 607722), ERC Advanced Grant (PROTEUS) has received funding from the European Research
vii2 | Molecular Analysis for Personalised Therapy
Council under the European Union’s Horizon 2020 research and innovation programme (grant agreement No. 835297), Chromavision – this project has received funding from the European’s Union Horizon 2020 research and innovation programme (grant agreement No. 665233). Disclosure: D. Biswas: Non-remunerated activity/ies, patent co-inventor: CRUK; Studentship: Jean Shanks Foundation MBPhD; Funding (self): MBPhD programme at University College London; Funding (institution): the NIHR BRC at University College London Hospitals. N.J. Birkbak: Nonremunerated activity/ies, patent co-inventor: CRUK; Fellow: Lundbeck Foundation; Funding (self): the Aarhus University Research Foundation; Funding (institution): Danish Cancer Society. N. McGranahan: Non-remunerated activity/ies, patent co-inventor: CRUK. Fellow: Sir Henry Dale Fellow, jointly funded by the Wellcome Trust and the Royal Society (Grant Number 211179/Z/18/Z); Funding (institution): CRUK, Rosetrees, the NIHR BRC at University College London Hospitals. C. Swanton: Advisory / Consultancy, Research grant / Funding (institution): Pfizer; Advisory / Consultancy, Research grant / Funding (institution): AstraZeneca; Advisory / Consultancy, Research grant / Funding (institution): BMS; Advisory / Consultancy, Research grant / Funding (institution): Roche; Advisory / Consultancy, Research grant / Funding (institution): Ventana; Advisory / Consultancy: Novartis; Advisory / Consultancy: GlaxoSmithKline; Advisory / Consultancy: MSD; Advisory / Consultancy: Celgene; Advisory / Consultancy: Illumina; Advisory / Consultancy: Genentech; Advisory / Consultancy, Shareholder / Stockholder / Stock options: GRAIL; Advisory / Consultancy: Medicxi; Advisory / Consultancy: Sarah Cannon Research Institute; Advisory / Consultancy: Dynamo Therapeutics; Shareholder / Stockholder / Stock options: ApoGen Biotechnologies; Shareholder / Stockholder / Stock options: Epic Bioscience; Shareholder / Stockholder / Stock options: Achilles Therapeutics. Funding (institution) Cancer Research UK (TRACERx and CRUK Cancer Immunotherapy Catalyst Network), the CRUK Lung Cancer Centre of Excellence, Stand Up 2 Cancer (SU2C), the Rosetrees Trust, Butterfield and Stoneygate Trusts, NovoNordisk Foundation (ID16584), the Prostate Cancer Foundation, the Breast Cancer Research Foundation (BCRF), National Institute for Health Research, the University College London Hospitals Biomedical Research Centre, and the Cancer Research UK University College London Experimental Cancer Medicine Centre.
6O
Comprehensive genomic profiling and outcomes among metastatic melanoma patients (pts) treated with first-line cancer immunotherapy (CIT) in a real-world setting
N. Sadetsky1, P. Lambert2, C. Julian2, J. Chen2, Y. Yan1 Genentech Inc., Roche USA, South San Francisco, CA, USA, 2Genentech Inc., South San Francisco, CA, USA
1
Background: Distinctive genomic subtypes (BRAF-mutated [BRAFmut], NRASmut, NF1mut, and triple wild type [wt]) have been identified in melanoma, but little is known about their distribution and association with outcomes outside of clinical trials. By linking longitudinal electronic health records (EHR) and comprehensive genomic profiling we aim to describe characteristics and outcomes among CIT-treated pts in a real-world setting. Methods: Metastatic melanoma pts in the de-identified Flatiron Health (FH)Foundation Medicine (FMI) Clinico-Genomic Database (CGDB), in which EHRderived data from FH are linked to genomic data from FMI, who received first-line pembrolizumab (pembro), nivolumab (nivo), ipilimumab (ipi), or ipi þ nivo between Jan 1, 2011, and Nov 30, 2018, were analyzed. Median OS was calculated from the start of first-line therapy and was evaluated according to BRAF status and genomic subtypes. Results: Of 656 melanoma pts in CGDB, 236 received first-line CIT. The majority of pts (69%, n ¼ 165) were BRAFwt and the rest (31%, n ¼ 71) were BRAFmut. Median age was 62.9 yrs (57.3 yrs for BRAFmut pts and 65.6 yrs for BRAFwt pts), 89% were white, and 97% were treated in community clinics. Among BRAFwt pts, 40% (n ¼ 66) were NRASmut, 32% (n ¼ 53) were NF1mut, and 28% (n ¼ 46) were triple wt. Pembro was used in 36% of pts, followed by ipi þ nivo (25%), nivo (20%), and ipi (18%). Almost half (47%) of the pts received subsequent therapy; of these pts, 53% (n ¼ 59) received CIT, 23% (n ¼ 26) received targeted therapy, and 24% (n ¼ 27) received other therapies. Median OS for BRAFwt and BRAFmut pts after first-line CIT was 38.6 mo (95% CI 20.2–not estimable [NE]) and 28.9 mo (95% CI 23.3–NE), respectively. Among BRAFwt pts, median OS was 44.9 mo (95% CI 28.9–NE) for NRASmut pts, 27.1 mo (95% CI 19.4–NE) for NF1mut pts, and 19.8 mo (95% CI 11.8-NE) for triple wt pts. Conclusions: Our study demonstrated that outcomes among melanoma pts receiving first-line CIT in real-world setting vary based on genomic subtypes and revealed, for the first time, differences in outcomes of major genomic subtypes in BRAFwt pts. Continued investigation of the association between specific genomic subtypes and survival in a real-world setting is needed. Editorial acknowledgement: Melanie Sweetlove, MSc (ApotheCom, Yardley, PA, USA). Legal entity responsible for the study: Genentech, a member of the Roche Group. Funding: Genentech, a member of the Roche Group. Disclosure: N. Sadetsky: Shareholder / Stockholder / Stock options, Full / Part-time employment: Genentech, a member of the Roche Group. P. Lambert: Shareholder / Stockholder / Stock options, Full / Part-time employment: Genentech, a member of the Roche Group. C. Julian: Shareholder / Stockholder / Stock options, Full / Part-time employment: Genentech, a member of the Roche Group. J. Chen: Shareholder / Stockholder / Stock options, Full / Part-time employment: Genentech, a member of the Roche Group. Y. Yan: Shareholder / Stockholder / Stock options, Full / Part-time employment: Genentech, a member of the Roche Group.
Volume 30 | Supplement 7 | November 2019
Downloaded from https://academic.oup.com/annonc/article-abstract/30/Supplement_7/mdz413.011/5613278 by guest on 24 December 2019
Background: Pancreatic Neuroendocrine Tumours (PanNETs) are the second most common form of pancreatic cancer. We previously identified two principal subtypes of PanNET: insulinomas (islet tumors; IT) and metastasis-like primaries (MLP), corresponding to the low- and high-grade classification of human PanNETs. This study describes the mechanisms by which PanNETs progress from IT to more aggressive MLP tumors and eventually metastasize. Methods: We profiled single cell transcriptomes, bulk mRNA and miRNA transcriptomes and the proteomes of primary and metastasis specimens from the genetically engineered mouse model of PanNETs (Rip1 Tag2). Results: We found that the IT tumours maintained the expression of mature ß cell markers. In contrast, the MLP tumours expressed pancreatic progenitor markers. By integrating the data on ß-cell differentiation from pancreatic progenitors to mature ß cells, we demonstrated that the tumour progression from IT to MLP follows the reverse embryonic and postnatal developmental path. Furthermore, we identified Hmgb3 and miR-181cd cluster as the MLP master regulators. Over-expression of the miR-181cd cluster in IT-like cancer cell-lines resulted in the acquisition of the MLP gene signature and MLP morphologic phenotypes, in addition to activation Hmgb3 expression. This suggested a central role for Hmgb3 in initiating the MLP program. Furthermore, inhibiting the expression of Hmgb3 in MLP-like cancer cell-lines resulted in a significant growth decrease, demonstrating the importance of Hmgb3 in the maintenance of MLP-like cell state in vitro. Using transcriptomic data from human patients, we evaluated the relevance of the MLP program in human tumours. We established that aggressive human PanNET tumours also follow the same reverse developmental trajectory of dedifferentiation. Notably, patients with high MLP genes expression had a worse prognosis. Conclusions: These data demonstrate dedifferentiation as a mechanism by which malignant neuroendocrine cancer cells acquire progenitor-like features, enabling them to become more aggressive and metastatic. In addition, miR-181cd cluster and Hmgb3 act as the core regulators in the initiation of dedifferentiation and maintenance of progenitor-like features in tumour cells. Legal entity responsible for the study: CMSO Group at ISREC, EPFL. Funding: Swiss National Science Foundation. Disclosure: All authors have declared no conflicts of interest.
Annals of Oncology