Anna Karlsson1, Hans Brunnström2, Patrick Micke3, Srinivas Veerla1, Johanna Mattsson3, Linnea La Fleur3, Johan Botling3, Mats Jönsson1, Christel Reuterswärd1, Maria Planck4, Johan Staaf5. 1. Division of Oncology and Pathology, Department of Clinical Sciences Lund, Lund University, Lund, Sweden. 2. Division of Oncology and Pathology, Department of Clinical Sciences Lund, Lund University, Lund, Sweden; Department of Pathology, Regional Laboratories Region Skåne, Lund, Sweden. 3. Department of Immunology, Genetics and Pathology, Uppsala University, Uppsala, Sweden. 4. Division of Oncology and Pathology, Department of Clinical Sciences Lund, Lund University, Lund, Sweden; Department of Respiratory Medicine and Allergology, Skåne University Hospital, Lund, Sweden. 5. Division of Oncology and Pathology, Department of Clinical Sciences Lund, Lund University, Lund, Sweden. Electronic address: johan.staaf@med.lu.se.
Abstract
INTRODUCTION: Large cell lung cancer (LCLC) and large cell neuroendocrine carcinoma (LCNEC) constitute a small proportion of NSCLC. The WHO 2015 classification guidelines changed the definition of the debated histological subtype LCLC to be based on immunomarkers for adenocarcinoma and squamous cancer. We sought to determine whether these new guidelines also translate into the transcriptional landscape of lung cancer, and LCLC specifically. METHODS: Gene expression profiling was performed by using Illumina V4 HT12 microarrays (Illumina, San Diego, CA) on samples from 159 cases (comprising all histological subtypes, including 10 classified as LCLC WHO 2015 and 14 classified as LCNEC according to the WHO 2015 guidelines), with complimentary mutational and immunohistochemical data. Derived transcriptional phenotypes were validated in 199 independent tumors, including six WHO 2015 LCLCs and five LCNECs. RESULTS: Unsupervised analysis of gene expression data identified a phenotype comprising 90% of WHO 2015 LCLC tumors, with characteristics of poorly differentiated proliferative cancer, a 90% tumor protein p53 gene (TP53) mutation rate, and lack of well-known NSCLC oncogene driver alterations. Validation in independent data confirmed aggregation of WHO 2015 LCLCs in the specific phenotype. For LCNEC tumors, the unsupervised gene expression analysis suggested two different transcriptional patterns corresponding to a proposed genetic division of LCNEC tumors into SCLC-like and NSCLC-like cancer on the basis of TP53 and retinoblastoma 1 gene (RB1) alteration patterns. CONCLUSIONS: Refined classification of LCLC has implications for diagnosis, prognostics, and therapy decisions. Our molecular analyses support the WHO 2015 classification of LCLC and LCNEC tumors, which herein follow different tumorigenic paths and can accordingly be stratified into different transcriptional subgroups, thus linking diagnostic immunohistochemical staining-driven classification with the transcriptional landscape of lung cancer.
INTRODUCTION: Large cell lung cancer (LCLC) and large cell neuroendocrine carcinoma (LCNEC) constitute a small proportion of NSCLC. The WHO 2015 classification guidelines changed the definition of the debated histological subtype LCLC to be based on immunomarkers for adenocarcinoma and squamous cancer. We sought to determine whether these new guidelines also translate into the transcriptional landscape of lung cancer, and LCLC specifically. METHODS: Gene expression profiling was performed by using Illumina V4 HT12 microarrays (Illumina, San Diego, CA) on samples from 159 cases (comprising all histological subtypes, including 10 classified as LCLC WHO 2015 and 14 classified as LCNEC according to the WHO 2015 guidelines), with complimentary mutational and immunohistochemical data. Derived transcriptional phenotypes were validated in 199 independent tumors, including six WHO 2015 LCLCs and five LCNECs. RESULTS: Unsupervised analysis of gene expression data identified a phenotype comprising 90% of WHO 2015 LCLC tumors, with characteristics of poorly differentiated proliferative cancer, a 90% tumor protein p53 gene (TP53) mutation rate, and lack of well-known NSCLC oncogene driver alterations. Validation in independent data confirmed aggregation of WHO 2015 LCLCs in the specific phenotype. For LCNEC tumors, the unsupervised gene expression analysis suggested two different transcriptional patterns corresponding to a proposed genetic division of LCNEC tumors into SCLC-like and NSCLC-like cancer on the basis of TP53 and retinoblastoma 1 gene (RB1) alteration patterns. CONCLUSIONS: Refined classification of LCLC has implications for diagnosis, prognostics, and therapy decisions. Our molecular analyses support the WHO 2015 classification of LCLC and LCNEC tumors, which herein follow different tumorigenic paths and can accordingly be stratified into different transcriptional subgroups, thus linking diagnostic immunohistochemical staining-driven classification with the transcriptional landscape of lung cancer.
Authors: Helena Cirenajwis; Martin Lauss; Maria Planck; Johan Vallon-Christersson; Johan Staaf Journal: Brief Bioinform Date: 2020-03-23 Impact factor: 11.622
Authors: Anna Karlsson; Helena Cirenajwis; Kajsa Ericson-Lindquist; Hans Brunnström; Christel Reuterswärd; Mats Jönsson; Cristian Ortiz-Villalón; Aziz Hussein; Bengt Bergman; Anders Vikström; Nastaran Monsef; Eva Branden; Hirsh Koyi; Luigi de Petris; Patrick Micke; Annika Patthey; Annelie F Behndig; Mikael Johansson; Maria Planck; Johan Staaf Journal: Sci Rep Date: 2019-03-26 Impact factor: 4.379
Authors: Natasha Rekhtman; Catherine M Pietanza; Joshua Sabari; Joseph Montecalvo; Hangjun Wang; Omar Habeeb; Kyuichi Kadota; Prasad Adusumilli; Charles M Rudin; Marc Ladanyi; William D Travis; Philippe Joubert Journal: Mod Pathol Date: 2017-09-08 Impact factor: 7.842