Advancing sarcoma diagnostics with expanded DNA methylation-based classification

Advancing sarcoma diagnostics with expanded DNA methylation-based classification | medRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-P4HH5NV'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search Advancing sarcoma diagnostics with expanded DNA methylation-based classification Natalie Jäger , David E. Reuss , Martin Sill , Daniel Schrimpf , Abigail K. Suwala , Philipp Sievers , Rouzbeh Banan , Felix Hinz , Ramin Rahmanzade , Henry Bogumil , Kaan Fuat Aras , Areeba Patel , Andrey Korshunov , Melanie Bewerunge-Hudler , Arjen HG. Cleven , Manel Esteller , Hanno Glimm , Wolfgang Hartmann , Simon Kreutzfeld , Christoph Heilig , Till Milde , Iver Petersen , Christian M. Vokuhl , Wolfgang Wick , Olaf Witt , Thibault Kervarrec , Evelina Miele , Jonathan Serrano , Stephan Frank , Karl Kashofer , Anne Mc Leer , Elke Pfaff , Melanie Pages , Arnault Tauziede-Espariat , Ferdinand Toberer , Henning B. Boldt , Petr Martinek , Sebastian Brandner , Mayara Euzebio , Aurore Siegfried , Jane Chalker , Patrik Harter , Romain Appay , Wolfgang Dietmaier , Martin Hasselblatt , Uta E. Flucke , Laura S. Hiemcke-Jiwa , David Solomon , Clara Frydrychowicz , Pascale Varlet , Benjamin Goeppert , Michaela Nathrath , Claudia Blattmann , Monika Sparber-Sauer , August Kolb , Michel Mittelbronn , Thomas Mentzel , Sandra Leisz , Anja Harder , Till Acker , Drew Pratt , Eva Wardelmann , Jamal Benhamida , Mark Ladanyi , Philipp Jurmeister , William Foulkes , Pamela Ajuyah , David Z. Ziegler , Jürgen Hench , Maikel JL. Nederkoorn , Yvonne MH. Versleijen-Jonkers , Gunhild Mechtersheimer , Sandro Krieg , Manfred Gessler , Daniel Baumhoer , Sam Behjati , Luca Bertero , Klaus Griwank , Dirk Schadendorf , Pancras CW. Hogendoorn , Jean-Francois Emile , Paul G. Kemps , Armin Jarosch , Michael W. Ronellenfitsch , Toni Su Idler , Daniela Aust , Sylvia Herold , Jessica Pablik , Maysa Al-Hussaini , Zied Abdullaev , Maximus Yeung , Marco Wachtel , Eva Brack , Felix KF. Kommoss , Markku Miettinen , Ken Aldape , Adrienne MH. Flanagan , Uta Dirksen , Kristian Pajtler , Thomas GP. Grünewald , Daniel Lipka , Stefan Fröhling , Christian Koelsche , Matija Snuderl , David Capper , Stefan M. Pfister , David TW. Jones , Felix Sahm , Andreas von Deimling doi: https://doi.org/10.1101/2025.06.30.25330543 Natalie Jäger 1 Division of Pediatric Neurooncology, German Cancer Research Center (DKFZ) and German Consortium for Translational Cancer Research (DKTK) , Heidelberg, Germany 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: andreas.vondeimling{at}med.uni-heidelberg.de David E. Reuss 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Martin Sill 1 Division of Pediatric Neurooncology, German Cancer Research Center (DKFZ) and German Consortium for Translational Cancer Research (DKTK) , Heidelberg, Germany 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Daniel Schrimpf 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Abigail K. Suwala 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Philipp Sievers 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Rouzbeh Banan 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Felix Hinz 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Ramin Rahmanzade 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Henry Bogumil 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Kaan Fuat Aras 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Areeba Patel 1 Division of Pediatric Neurooncology, German Cancer Research Center (DKFZ) and German Consortium for Translational Cancer Research (DKTK) , Heidelberg, Germany 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Andrey Korshunov 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany 6 National Center for Tumor Diseases Heidelberg (NCT) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Melanie Bewerunge-Hudler 7 Genomics and Proteomics Core Facility, German Cancer Research Center (DKFZ) , 69120, Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Arjen HG. Cleven 8 Department of Pathology and Medical Biology, University Medical Center Groningen , Groningen, the Netherlands 9 Department of Pathology, Amsterdam University Medical Centers , Amsterdam, the Netherlands Find this author on Google Scholar Find this author on PubMed Search for this author on this site Manel Esteller 10 Institucio Catalana de Recerca i Estudis Avançats (ICREA) , Barcelona, Catalonia, Spain 11 Physiological Sciences Department, School of Medicine and Health Sciences, University of Barcelona (UB) , Barcelona, Catalonia, Spain 12 Centro de Investigacion Biomedica en Red Cancer (CIBERONC) , Madrid, Spain 13 Cancer Epigenetics Group, Josep Carreras Leukaemia Research Institute (IJC) , Badalona, Barcelona, Catalonia, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site Hanno Glimm 14 German Cancer Consortium (DKTK) , Dresden, Germany 15 Translational Functional Cancer Genomics, DKFZ , Heidelberg, Germany 16 Translational Medical Oncology, National Center for Tumor Diseases Dresden (NCT/UCC), Technical University Dresden and Helmholtz-Zentrum Dresden 17 Translational Medical Oncology, Technical University Dresden , 01307, Dresden, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Wolfgang Hartmann 18 Gerhard-Domagk-Institute of Pathology, University of Münster , Münster, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Simon Kreutzfeld 19 Division of Translational Medical Oncology, German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Christoph Heilig 19 Division of Translational Medical Oncology, German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Till Milde 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany 6 National Center for Tumor Diseases Heidelberg (NCT) , Heidelberg, Germany 20 Clinical Cooperation Unit Pediatric Oncology, German Cancer Research Center (DKFZ) and German Cancer Consortium (DKTK) , Heidelberg, Germany 21 Department of Pediatric Oncology, Hematology, Immunology and Pulmonology, Heidelberg University Hospital , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Iver Petersen 22 Institute of Pathology , Waldklinikum Gera, Gera, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Christian M. Vokuhl 23 Department of Pathology, Section Paidopathology, University Hospital Bonn , 53127 Bonn, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Wolfgang Wick 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany 6 National Center for Tumor Diseases Heidelberg (NCT) , Heidelberg, Germany 24 Department of Neurology, Heidelberg University Hospital, Heidelberg University , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Olaf Witt 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany 6 National Center for Tumor Diseases Heidelberg (NCT) , Heidelberg, Germany 20 Clinical Cooperation Unit Pediatric Oncology, German Cancer Research Center (DKFZ) and German Cancer Consortium (DKTK) , Heidelberg, Germany 21 Department of Pediatric Oncology, Hematology, Immunology and Pulmonology, Heidelberg University Hospital , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Thibault Kervarrec 25 Department of Pathology, Centre Hospitalier Universitaire de Tours, Université de Tours , Tours, France 26 CARADERM Network , Lille, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Evelina Miele 27 Onco-Hematology, Cell Therapy, Gene Therapies and Hemopoietic Transplant, Bambino Gesù Children’s Hospital , IRCCS, Rome, Italy Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jonathan Serrano 28 Department of Pathology, NYU Langone Health and NYU Grossman School of Medicine , New York, New York, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Stephan Frank 29 Institute for Medical Genetics and Pathology, University Hospital Basel , Basel, Switzerland Find this author on Google Scholar Find this author on PubMed Search for this author on this site Karl Kashofer 30 Department of Obstetrics & Gynecology and Diagnostic-und Research Institute of Pathology, Medical University of Graz , Austria Find this author on Google Scholar Find this author on PubMed Search for this author on this site Anne Mc Leer 31 Grenoble Alpes University Hospital Pathology Department, Institute for Advanced Biosciences UGA/INSERM U1209/CNRS 5309, Grenoble Alpes University , Grenoble, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Elke Pfaff 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany 21 Department of Pediatric Oncology, Hematology, Immunology and Pulmonology, Heidelberg University Hospital , Heidelberg, Germany 32 Division of Pediatric Glioma Research, German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Melanie Pages 33 GHU-Paris - Sainte-Anne Hospital, Department of Neuropathology, Paris University , Paris, France 34 Institut Curie, Paris Sciences Lettres University , SIREDO, INSERM U830, Laboratory of Translational Research in Paediatric Oncology , Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Arnault Tauziede-Espariat 35 Department of Neuropathology, GHU Paris - Psychiatry and Neuroscience, Sainte-Anne Hospital , Paris, 75014, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Ferdinand Toberer 36 Department of Dermatology, University of Heidelberg , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Henning B. Boldt 37 Department of Clinical Research and BRIDGE, University of Southern Denmark , 5000 Odense, Denmark 38 Department of Pathology, Odense University Hospital , 5000 Odense, Denmark Find this author on Google Scholar Find this author on PubMed Search for this author on this site Petr Martinek 39 Bioptical Laboratory, Ltd , Pilsen, Czech Republic Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sebastian Brandner 40 Division of Neuropathology, University College London Hospitals NHS Foundation Trust , Queen Square, London WC1N 3BG, UK 41 Department of Neurodegenerative Diseases, Queen Square Institute of Neurology, University College London , London WC1N 3BG, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mayara Euzebio 42 Genetics and Molecular Biology, Institute of Biology, State University of Campinas , Campinas 13083-862, SP, Brazil 43 Research Center, Boldrini Children’s Hospital , Campinas 13083-884, SP, Brazil Find this author on Google Scholar Find this author on PubMed Search for this author on this site Aurore Siegfried 44 Pathology Department, Toulouse University Hospital , Toulouse, France 45 UFR Santé, Université de Toulouse , Toulouse, France 46 Toulouse University , Inserm, CNRS, Centre de Recherches en Cancérologie de Toulouse , Toulouse, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jane Chalker 47 Specialist Integrated Haematology and Malignancy Diagnostic Service-Acquired Genomics, Great Ormond Street Hospital for Children NHS Foundation Trust , London, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site Patrik Harter 48 Institute of Neuropathology, Faculty of Medicine , LMU Munich, Munich, Germany 49 Neurological Institute (Edinger Institute), University Hospital Frankfurt , Frankfurt am Main, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Romain Appay 50 Department of Pathological Anatomy and Neuropathology, Timone Hospital , APHM, Marseille, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Wolfgang Dietmaier 51 Institute of Pathology, Center for Molecular Pathology Diagnosis, University of Regensburg , Regensburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Martin Hasselblatt 52 Institute of Neuropathology, University Hospital Münster , Pottkamp 2, 48149, Münster, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Uta E. Flucke 53 Diagnostic Laboratory, Princess Maxima Center for Pediatric Oncology , Utrecht, The Netherlands 54 Department of Pathology, Radboud University Medical Center , Nijmegen, The Netherlands Find this author on Google Scholar Find this author on PubMed Search for this author on this site Laura S. Hiemcke-Jiwa 55 Princess Maxima Center and UMC Utrecht , The Netherlands Find this author on Google Scholar Find this author on PubMed Search for this author on this site David Solomon 56 Neuropathology, Stanford University School of Medicine , Stanford, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Clara Frydrychowicz 57 Division of Neuropathology, University Hospital Leipzig , 04103, Leipzig, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Pascale Varlet 35 Department of Neuropathology, GHU Paris - Psychiatry and Neuroscience, Sainte-Anne Hospital , Paris, 75014, France 58 Institut de Psychiatrie et Neurosciences de Paris (IPNP) , UMR S1266, INSERM, IMA-BRAIN, Paris, France 59 Laboratory of Somatic Genetics, Université de Paris, France3 Curie Institute Hospital , Paris, Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Benjamin Goeppert 60 Institute of Pathology and Neuropathology , RKH Klinikum Ludwigsburg, Ludwigsburg, Germany 61 Department of Pathology, Heidelberg University Hospital , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Michaela Nathrath 62 Children’s Cancer Research Centre and Department of Pediatrics , Klinikum rechts der Isar, Technische Universität München , Munich, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Claudia Blattmann 63 Klinikum Stuttgart - Olgahospital , Zentrum für Kinder-, Jugend-und Frauenmedizin, Pädiatrie, Stuttgart, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Monika Sparber-Sauer 64 Klinikum der Landeshauptstadt Stuttgart gKAöR, Olgahospital, Stuttgart Cancer Center , Zentrum für Kinder-, Jugend-und Frauenmedizin, Pädiatrie 5 (Pädiatrische Onkologie, Hämatologie, Immunologie), Stuttgart, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site August Kolb 65 Department of Pathology, New York University Langone Health and School of Medicine , New York, New York, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Michel Mittelbronn 66 Department of Cancer Research (DoCR), Luxembourg Institute of Health (LIH) , 1210 Luxembourg , Luxembourg 67 Department of Life Sciences and Medicine (DLSM), University of Luxembourg , 4365 Esch-sur- Alzette, Luxembourg 68 Luxembourg Center of Neuropathology (LCNP) , 3555 Dudelange, Luxembourg 69 Luxembourg Centre for Systems Biomedicine (LCSB), University of Luxembourg (UL) , 4362 Esch- sur-Alzette, Luxembourg Find this author on Google Scholar Find this author on PubMed Search for this author on this site Thomas Mentzel 70 Dermatopathology Friedrichshafen , Friedrichshafen, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sandra Leisz 71 Department of Neurosurgery, University Medicine Halle , Halle, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Anja Harder 72 Cure NF Research Group, Anatomy and Cell Biology, University Medicine Halle , Halle, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Till Acker 73 Institute of Neuropathology, Justus-Liebig University Giessen , Giessen, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Drew Pratt 74 Laboratory of Pathology, National Cancer Institute , NIH, Bethesda, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Eva Wardelmann 18 Gerhard-Domagk-Institute of Pathology, University of Münster , Münster, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jamal Benhamida 75 Department of Pathology and Laboratory Medicine, Memorial Sloan Kettering Cancer Center , New York City, New York, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mark Ladanyi 75 Department of Pathology and Laboratory Medicine, Memorial Sloan Kettering Cancer Center , New York City, New York, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Philipp Jurmeister 76 Institute of Pathology, Ludwig Maximilians University Hospital Munich , 80336 Munich, Germany 77 German Cancer Consortium (DKTK) , Munich, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site William Foulkes 78 Department of Human Genetics, McGill University , Montreal, Quebec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site Pamela Ajuyah 79 Children’s Cancer Institute, Lowy Cancer Research Centre , UNSW, Kensington, NSW, Australia Find this author on Google Scholar Find this author on PubMed Search for this author on this site David Z. Ziegler 80 Kids Cancer Centre, Sydney Children’s Hospital , Randwick, NSW, Australia 81 Children’s Cancer Institute, Lowy Cancer Centre , UNSW Sydney, Kensington, NSW, Australia 82 School of Clinical Medicine, UNSW Medicine & Health, UNSW Sydney , Kensington, NSW, Australia Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jürgen Hench 29 Institute for Medical Genetics and Pathology, University Hospital Basel , Basel, Switzerland Find this author on Google Scholar Find this author on PubMed Search for this author on this site Maikel JL. Nederkoorn 83 Department of Medical Oncology, Radboud University Medical Center , Nijmegen, The Netherlands Find this author on Google Scholar Find this author on PubMed Search for this author on this site Yvonne MH. Versleijen-Jonkers 84 Department of Medical Oncology, Radboud University Medical Centre , 6525 GA Nijmegen, The Netherlands Find this author on Google Scholar Find this author on PubMed Search for this author on this site Gunhild Mechtersheimer 61 Department of Pathology, Heidelberg University Hospital , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sandro Krieg 85 Department of Neurosurgery, University Hospital Heidelberg, Heidelberg University , 69117 Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Manfred Gessler 86 Theodor-Boveri-Institute/Biocenter, Developmental Biochemistry, University of Wuerzburg , Wuerzburg, Germany 87 Comprehensive Cancer Center Mainfranken, University of Wuerzburg , Wuerzburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Daniel Baumhoer 29 Institute for Medical Genetics and Pathology, University Hospital Basel , Basel, Switzerland Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sam Behjati 88 Department of Paediatrics, University of Cambridge; Cambridge , UK 89 Wellcome Sanger Institute; Hinxton , CB10 1SA, UK 90 Cambridge University Hospitals NHS Foundation Trust; Cambridge , CB2 0QQ, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site Luca Bertero 91 Pathology Unit, Department of Laboratory Medicine, Department of Medical Sciences, University of Turin , Turin, Italy Find this author on Google Scholar Find this author on PubMed Search for this author on this site Klaus Griwank 92 Department of Dermatology, University Hospital Essen, University of Duisburg-Essen , Germany & German Cancer Consortium (Deutsches Konsortium für Translationale Krebsforschung, DKTK), Essen, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Dirk Schadendorf 92 Department of Dermatology, University Hospital Essen, University of Duisburg-Essen , Germany & German Cancer Consortium (Deutsches Konsortium für Translationale Krebsforschung, DKTK), Essen, Germany 93 Comprehensive Cancer Center (Westdeutsches Tumorzentrum), University Hospital Essen, Essen & National Center for Tumor Diseases (NCT) West , Essen, Germany 94 Research Center One Health, University Duisburg-Essen , Essen, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Pancras CW. Hogendoorn 95 Department of Pathology, Leiden University Medical Center , Leiden, The Netherlands Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jean-Francois Emile 96 Paris-Saclay University, Versailles SQY University , EA4340-BECCOH, Assistance Publique– Hôpitaux de Paris (AP-HP), Ambroise-Paré Hospital , Smart Imaging, Service de Pathologie, Boulogne, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Paul G. Kemps 97 Department of Pathology, Leiden University Medical Center , Leiden, the Netherlands Find this author on Google Scholar Find this author on PubMed Search for this author on this site Armin Jarosch 98 Institute of Pathology, Charité - Universitätsmedizin Berlin , Berlin, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Michael W. Ronellenfitsch 99 Senckenberg Institute of Neurooncology, University Hospital Frankfurt , Frankfurt, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Toni Su Idler 100 Institute of Pathology, Technical University Dresden , 01307, Dresden, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Daniela Aust 100 Institute of Pathology, Technical University Dresden , 01307, Dresden, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sylvia Herold 100 Institute of Pathology, Technical University Dresden , 01307, Dresden, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jessica Pablik 100 Institute of Pathology, Technical University Dresden , 01307, Dresden, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Maysa Al-Hussaini 101 Department of Pathology and Laboratory Medicine, King Hussein Cancer Center , Amman, Jordan Find this author on Google Scholar Find this author on PubMed Search for this author on this site Zied Abdullaev 74 Laboratory of Pathology, National Cancer Institute , NIH, Bethesda, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Maximus Yeung 102 Department of Pathology, The University of Hong Kong, Queen Mary Hospital , Hong Kong Find this author on Google Scholar Find this author on PubMed Search for this author on this site Marco Wachtel 103 University Children’s Hospital , Oncology, Balgrist Campus, Zurich, Switzerland Find this author on Google Scholar Find this author on PubMed Search for this author on this site Eva Brack 104 Division of Pediatric Hematology/Oncology, Department of Pediatrics , Inselspital, Bern University Hospital, University of Bern , Bern, Switzerland Find this author on Google Scholar Find this author on PubMed Search for this author on this site Felix KF. Kommoss 61 Department of Pathology, Heidelberg University Hospital , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Markku Miettinen 74 Laboratory of Pathology, National Cancer Institute , NIH, Bethesda, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Ken Aldape 74 Laboratory of Pathology, National Cancer Institute , NIH, Bethesda, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site Adrienne MH. Flanagan 105 The Royal National Orthopaedic Hospital Trust , Brockley Hill, Stanmore HA7 4LP, UK 106 UCL Cancer Institute , London WC1E 6DD, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site Uta Dirksen 107 Pediatric Oncology & Hematology, Pediatrics III, University Hospital of Essen , Essen, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Kristian Pajtler 108 Division of Pediatric Neurooncology, Hopp Children’s Cancer Center Heidelberg (KiTZ); German Cancer Research Center Heidelberg (DKFZ) and Heidelberg University Hospital, Heidelberg; National Center for Tumor Diseases (NCT) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Thomas GP. Grünewald 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany 6 National Center for Tumor Diseases Heidelberg (NCT) , Heidelberg, Germany 61 Department of Pathology, Heidelberg University Hospital , Heidelberg, Germany 109 Division of Translational Pediatric Sarcoma Research, German Cancer Research Center (DKFZ) , German Cancer Consortium (DKTK), 69120 Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Daniel Lipka 19 Division of Translational Medical Oncology, German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Stefan Fröhling 6 National Center for Tumor Diseases Heidelberg (NCT) , Heidelberg, Germany 19 Division of Translational Medical Oncology, German Cancer Research Center (DKFZ) , Heidelberg, Germany 110 German Cancer Consortium (DKTK) , Heidelberg, Germany 111 Institute of Human Genetics, Heidelberg University , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Christian Koelsche 76 Institute of Pathology, Ludwig Maximilians University Hospital Munich , 80336 Munich, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Matija Snuderl 65 Department of Pathology, New York University Langone Health and School of Medicine , New York, New York, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site David Capper 112 Department of Neuropathology, Charité - Universitätsmedizin Berlin , Berlin, Germany 113 German Cancer Consortium (DKTK) , Berlin, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Stefan M. Pfister 1 Division of Pediatric Neurooncology, German Cancer Research Center (DKFZ) and German Consortium for Translational Cancer Research (DKTK) , Heidelberg, Germany 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany 5 Department of Pediatric Hematology and Oncology, Heidelberg University Hospital , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site David TW. Jones 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany 114 Pediatric Glioma Research Group, German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Felix Sahm 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Andreas von Deimling 2 Hopp Children’s Cancer Center Heidelberg (KiTZ) , Heidelberg, Germany 3 Department of Neuropathology, Institute of Pathology, University Hospital Heidelberg , Heidelberg, Germany 4 Clinical Cooperation Unit Neuropathology , German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ) , Heidelberg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Abstract Full Text Info/History Metrics Supplementary material Data/Code Preview PDF Abstract Purpose Sarcomas pose a severe diagnostic challenge. A wide variety of these distinct entities need to be distinguished from each other and from less aggressive types of mesenchymal tumors, to ensure correct clinical management. A machine learning based classifier for sarcomas utilizing DNA methylation data from 1077 tumors recognizing 62 sarcoma types has already been developed and termed the sarcoma classifier, which we published in 2021. Here we present a major advancement of the scale and precision of the sarcoma classifier. Methods DNA methylation profiles and histologic data from an unprecedented multi-institutional cohort of mesenchymal tumors were collected and analyzed. Utilizing a machine learning approach, the classifier was rigorously validated through five-fold nested cross-validation, achieving a 98% class-level accuracy and a Brier score of 0.017, indicative of well-calibrated probability estimates. Results The sarcoma classifier v13.1 was developed based on a training set of 4377 methylation profiles from sarcomas and less aggressive mesenchymal tumors comprising 116 tumor sub-classes and 4 control groups forming 93 distinct methylation classes. Performance was validated using four independent cohorts, comprising a total of 1547 mesenchymal tumors. A methylation-based classifier prediction was obtained in 73% of cases in the validation sets, of which 91% matched the original histopathology diagnosis, thereby increasing diagnostic confidence. The classifier enabled a definitive molecular diagnosis or tumor reclassification in 6% of cases with inconclusive or ambiguous histological findings. Conclusion Adding new sarcoma types and expanding tumor sample numbers in each methylation class in the new sarcoma classifier decisively increased the number of diagnostic predictions and improved match with histologic evaluation. This substantial advancement will promote clinical implementation of the tool for the diagnosis of mesenchymal tumor lesions. Introduction The challenges of diagnosing sarcomas result in a high interobserver variation, requiring of reevaluation by referral centers, and the performing of an increasing array of different molecular analyses 1 , 2 . Typical molecular analyses focus on DNA and RNA sequencing or FISH in order to find specific genetic alterations. Approximately half of the sarcoma types exhibit highly characteristic gene fusions, and therefore, cDNA sequencing is of utmost relevance in these tumors. Another molecular approach to sarcoma classification has previously been established using DNA methylation data, leveraging insights gained from the brain tumor classifier published in 2018 3 . Subsequently, a sarcoma classifier version v12.3 had been developed, which recognized 62 tumor methylation classes covering already a wide spectrum of mesenchymal tumors 4 . In contrast to sequencing approaches searching for pathognomonic mutations, the methylation-based approach aims at identifying methylation patterns which are closely related to the cell of origin of the tumor cells. This implies the need for a comprehensive sample reference set able to cope with the intertumoral variance of DNA methylation patterns. Expanding the reference dataset allows for broader representation of the biological diversity within and across tumor entities, which may facilitate the identification of additional methylation classes and increase the confidence and specificity of classifier predictions. To this end we have advanced the existing tool to the new sarcoma classifier v13.1 by adding 62 additional tumor types or subtypes. In addition, our hierarchical system leverages the well-calibrated classifier probability scores by aggregating subclass- level scores at higher tiers, thereby providing more robust diagnostic guidance at higher-order levels. A total of 120 subclasses is covered in 89 different tumor methylation classes and 4 control classes, based on a reference set containing 4377 DNA methylation array profiles ( Fig. 1 ). Download figure Open in new tab Figure 1. Comparison of the sarcoma classifier versions v12.3 and v13.1 (a) The size of the reference dataset was expanded from 1077 to 4377 samples. (b) The number of tumor methylation classes increased from 62 in v12.3 to 89 in v13.1, including additional control classes. (c) Sankey diagram to illustrate how the reference dataset from sarcomas and less aggressive mesenchymal tumors comprising 116 tumor and 4 control groups forms 93 distinct methylation classes. (d) Overview of tumor types and subtypes included in v13.1, highlighting refined subclass definitions, additions, and restructuring compared to v12.3. Notably, several entities were subclassified (e.g., AS, SFT, OFMT, GIST), while others were newly introduced (e.g., CMF, ABC, ADA). Material and Methods DNA-Methylation Array Processing The Illumina Infinium HumanMethylation450 (450k) array, Illumina Infinium MethylationEPIC (EPIC) array and Illumina Infinium MethylationEPICv2 (EPICv2) microarrays were used to obtain genome-wide DNA methylation data for tumor samples and normal control tissues according to the manufacturer’s instructions (Illumina, San Diego, USA). DNA methylation data were generated from both fresh-frozen and formalin-fixed paraffin-embedded (FFPE) tissue samples. For most of the fresh-frozen samples and FFPE tissue samples, between 50 ng and 500 ng of DNA was used as input material. On-chip quality metrics of all samples were carefully controlled. All computational analyses were performed in R version 4.3.3 (R Development Core Team, 2024). Raw signal intensities were obtained from IDAT files using the minfi Bioconductor package version 1.21.4 5 . Illumina EPIC, EPICv2 and 450k samples were merged into a combined dataset by selecting the intersection of probes present on all arrays (combineArrays function in minfi). Each sample was individually normalized by performing a background correction (shifting of the 5% percentile of negative control probe intensities to 0) and a dye-bias correction (scaling of the mean of normalization control probe intensities to 10,000) for both color channels. Subsequently, a correction for the type of material (FFPE/frozen) and array type (450k/EPIC(v2)) was performed by fitting univariable linear models to the log2-transformed intensity values (removeBatchEffect function in the limma package version 3.30.11). The methylated and unmethylated signals were corrected individually. Beta-values were calculated from the retransformed intensities using an offset of 100, as recommended by Illumina. Methylation Data Quality Control The so-called “detectionP” is based on detection p-values for all probed genomic positions ( https://rdrr.io/bioc/minfi/man/detectionP.html ). The detection p-value threshold indicates the probability that the observed signal is due to background noise. A cutoff of p < 0.05 is usually recommended for detection p-values. Minfi further provides a quality control metric that uses the log median intensity in both the methylated (M; Median Log2 Meth) and unmethylated (U; Median Log2 Unmeth) channels. When plotting these two medians against each other, good samples tend to cluster together, while failed samples tend to separate and have lower median intensities. The reference training data set as well as the validation data sets were filtered to contain only samples, which exceed the following three data quality-relevant control metrics: “Fraction of detectionP 0.95; Median Log2 Meth > 8; Median Log2 Unmeth > 8. Further, all samples of the data set indicating matching genotypes based on the SNPs on the methylation arrays were also de-duplicated to prevent having included the same patient sample more than once. Classifier training Classifier training was performed as described previously 3 , 6 . First, we applied a permutation-based variable importance measure (R-package randomForest v4.7-1.2) to select the 10,000 most informative CpG probes as features for the final Random Forest (RF). Unbalanced class sample sizes were considered by down sampling each bootstrap sample to the minority class. Next, a ridge-penalized multinomial logistic-regression model (R-package glmnet v4.1-8) was fitted to calibrate the RF output, mapping raw prediction scores to probability estimates. An optimal penalization parameter was chosen by a ten-fold cross-validation. Combining classifier outputs with a logistic regression model is an ensemble strategy known as stacking 51 . Classifier validation To evaluate the classifier, a five-fold nested cross-validation scheme generated out-of-sample RF scores that enabled us to fit and validate the calibration models in each fold. To measure the performance of the classifier the following metrics and figures were generated: Accuracy, Balanced Accuracy, F1, Matthews Correlation Coefficient (R-package mltest v1.0.1), Confusion Matrix, multiclass Log Loss, multiclass Brier Score, and Calibration Plots (R-package rms v6.8-2). In addition, receiver operating characteristics (ROC) curves and accompanying areas under the curve (AUC) were generated by binarizing multiclass predictions (correct vs. incorrect) (R-package pROC v1.18.5). Estimating tumor cell content from DNA methylation data Tumor cell content for all reference and validation samples was estimated using a deconvolution approach based on a previously published reference signature matrix developed for immune cell deconvolution in CNS tumors ( https://pubmed.ncbi.nlm.nih.gov/32859926 ). This matrix was derived from DNA methylation profiles of pure, flow-sorted cell populations, including profiles from cancer cell lines representative of CNS tumors and malignant rhabdoid tumors (MRT). In addition, for samples analysed with the HumanMethylation450 (450k) or Illumina Infinium MethylationEPIC (EPIC) array, the R package RF_Purify was used to estimate the tumor cell content (EPICv2 samples cannot be analysed with RF_Purify). The tumor cell content estimates from both methods were well correlated (Suppl. Fig 4). CNV analysis Gene amplifications and homozygous deletions (deep deletions) were called according to a published algorithm ( https://forum.depmap.org/t/defining-deep-deletions-and-amplifications/710/5 ). Values lower than -1.2 on a log scale were scored as homozygous deletions and values higher than 0.75 on a log scale were scored as amplification. Results Compiling the sarcoma classifier reference set The present dataset substantially expands upon the reference set of the established sarcoma classifier 4 . While v12.3 was generated on cases with morphological expert review and extensive molecular workup, the updated classifier version employs a combined approach, incorporating both molecularly well- annotated cases and cases selected based on matching diagnoses aligned with their association to specific methylation-defined clusters. The number of reference cases including controls increased from 1077 to 4377 cases ( Fig. 1a ), which have all been subjected to t-Distributed Stochastic Neighbor Embedding (t-SNE) and to Uniform Manifold Approximation and Projection (UMAP) in order to ensure their association with distinct methylation classes. Of the 62 tumor methylation classes from the sarcoma classifier v12.3 ( Fig. 1b ), 47 remained unchanged while 15 were modified either by subdivision or by combining with additional tumor methylation classes. The updated sarcoma reference set v13.1 now includes 55 new tumor classes not represented in the initial version ( Fig. 1d ). Sample numbers in distinct methylation classes ranged from 6 to 205 cases. Full descriptions matching to the abbreviations of methylation classes and information on case numbers within each methylation class is provided in supplementary table 1. Controls included, the updated reference set contains 93 methylation classes relating to 124 tumor and four control groups ( Figure 1b , c). The 128 tissue types contain 117 mesenchymal tumors, four controls and non-mesenchymal tumors including SCC (squamous cell carcinoma (cutaneous)), LCH (Langerhans cell histiocytosis), HS (histiocytic sarcoma), MEL (melanoma), MEL_MUC (melanoma mucosal) and MEL_DES (melanoma desmoplastic). The non- mesenchymal tumors were included, because they constitute relevant differential diagnoses to several sarcoma types. One entity, the SCCOHT (small cell carcinoma of the ovary, hypercalcemic type), was initially considered an epithelial tumor but more recently is discussed in the context of sarcomas or primitive germ cell tumors 7 . Brain specific sarcomas such as gliosarcoma, oligosarcoma, and primary intracranial sarcoma DICER1-mutant were not included, because they are represented in the organ specific brain tumor classifier v12.8 ( https://app.epignostix.com/#/classifiers ). Further including these sarcomas would have required to also include oligodendrogliomas, glioblastomas and other brain tumors into the sarcoma classifier v13.1 for clear separation of these tumor types. The methylation classes are briefly described in supplementary table 2. All 4377 tissue samples constituting the reference set have been analyzed for chromosomal alterations manifesting as gains and losses and for amplifications and homozygous deletions in a specific gene set. An overview of the CNV distribution in all tumor methylation classes is given in supplementary Fig. 2. To provide guidance on the available information on the new and existing classes and subclasses and their alignment with the current WHO classification, we introduced an evidence level annotation as specified in Fig.1d and Fig. 2a and supplementary table 2: Level a - Tumor type/subtype identical to WHO 2020; Level b - Large single or more than one smaller dataset published describing the type/subtype as molecularly and/or clinically distinct, or the methylation class represents a distinct fraction of an established WHO 2020 tumor class; Level c - Single small dataset or case series; Level d- solely based on clusters in tSNE/UMAP. The annotation is typically provided at the most granular layer, plus in part also at a higher layer if that instead matches a WHO type/subtype. Download figure Open in new tab Figure 2. Overview of the DNA methylation-based sarcoma reference cohort used for training the v13.1 sarcoma classifier. (a) Overview of 89 tumour and four control DNA methylation classes included in the v13.1 sarcoma classifier reference cohort. The sarcoma methylation classes are grouped according to the WHO 2020 scheme. Letters in rounded brackets before abbreviation of the class indicate ‘evidence level’ for the respective class. Full names and further details of the methylation classes are given in supplementary tables 1 and 2. (b) t-SNE projection of complete reference sample set (4377 DNA methylation profiles) used for training the v13.1 sarcoma classifier. Classifier v13.1 generation and cross validation performance Unsupervised clustering using t-SNE and UMAP, along with cross-validation results, revealed substantial overlap in the methylation profiles of a small number of tumor types and subtypes. Due to these overlapping methylation profiles, reliable separation of certain entities was not feasible. Therefore, seven combined methylation classes were defined by grouping these entities for classifier training, even though they do not represent established tumor subclasses. These combined methylation classes included only non-sarcoma tumors that are relevant in the differential diagnosis of sarcomas, such as melanomas (supplementary table 1). In summary, the v13.1 sarcoma classifier was trained using 116 subclasses and four groups for normal control tissue ( Fig. 1c ). The updated sarcoma classifier model was developed using a Random Forest–based approach 8 as previously described 3 , 4 . To evaluate its performance, the classifier was validated using a five-fold nested cross-validation scheme. All subclasses achieved a balanced accuracy greater than 0.64, with 101 out of 120 subclasses exceeding 0.9 in accuracy ( Fig. 3a ). The newly introduced hierarchical system leverages these well-calibrated probabilities by aggregating subclass-level scores at higher tiers, specifically for the subclasses with lower accuracy, thereby providing more robust diagnostic guidance at these higher-order tiers. Most of the more granular subclasses currently lack direct clinical implications, making aggregated probability scores at the higher Class level sufficient for guiding diagnostic decisions. Download figure Open in new tab Figure 3. Performance of the v13.1 sarcoma classifier in five-fold nested cross-validation. (a) Bar plot showing the balanced accuracy for each of the 93 subclasses, as derived from five-fold nested cross-validation. Subclass combinations into classes are indicated where applicable by a box with dashed lines. (b) Table summarizing overall performance metrics—accuracy, F1-score, log loss, and Brier score—evaluated at each hierarchical level (subclass, class). (c) Scatterplot of subclass- specific maximum specificity thresholds with color-coded subclasses and a red dashed line at 0.9 marking the recommended threshold for the calibrated score. (d) Calibration plot comparing predicted probabilities with observed outcomes, illustrating the degree of score calibration across subclasses. Overall, the classifier achieved a 98% Class-level accuracy with a Brier Score of 0.0168, indicating well- calibrated probability estimates ( Figure 3b ). As with the previous v12.3 version, we recommend a score threshold of 0.9 at the Class level as a “high- confidence prediction” for the v13.1 sarcoma classifier ( Fig. 3c, d ). This threshold offers a practical balance between clarity in communication and robust performance in clinical settings. To identify statistically optimal thresholds, we performed ROC analyses for each subclass in a one-against-all framework and computed, for each subclass, the threshold that maximizes the specificity ( Fig. 3c ). While the optimal thresholds ranged from as low as 0.13 up to 0.99—indicating that certain classes, such as adipocytic sarcomas, may require higher cutoffs for maximal specificity—the 0.9 threshold still ensures high specificity for the vast majority of classes (max. specificity mean of all subclasses = 0.6; Fig. 3c ). Methylation classes comprising several tumor entities The v13.1 reference set did not allow for clear separation of all tumor subclasses included, resulting in several methylation classes containing two or more tumor entities. Many of these methylation classes combined tumors assumed to be highly related to each other. Examples are methylation classes (mc) AFX/PDS (atypical fibroxanthoma/ pleomorphic dermal sarcoma) 9 , mc DF/ATFH (dermatofibroma/ (atypical fibrous histiocytoma), mc JXG/AXG (juvenile/adult xanthogranuloma), mc MESO (mesothelioma) and mc ALT/WDLS (atypical lipomatous tumor/ well differentiated liposarcoma), the latter separated on the basis of tumor location and potential respectability 10 with both tumor types having a high incidence of MDM2 amplifications. Several methylation classes contain lesions of different morphology or tumor localization but sharing similar driver mutations such as mc MLCT/MEL_UV (melanocytoma/ uveal melanoma) with GNAQ or GNA11 mutations, and mc ERMS_DICER1/PPB (embryonal rhabdomyosarcoma DICER1-mutant / pleuropulmonary blastoma) with DICER1 mutations. Common to tumors in mc NF/NF_PLEX/NST_ERBB2 (neurofibroma / neurofibroma plexiform / nerve sheath tumour, ERBB2-mutant) is growth in peripheral nerves and NF1 mutations in NF and NF_PLEX while NST_ERBB2 are characterized by mutations in the receptor tyrosine kinase gene ERBB2 . The mc MEL/MEL_MUC/MEL_DES (melanoma / melanoma-mucosal/melanoma demoplastic) comprises cutaneous, mucosal and desmoplastic melanoma. The mc FDY/OSF (fibrous dysplasia/ ossifying fibroma) combines two lower grade osseous lesions. Of these lesions, FDY typically carry GNAS mutations. The mc LMO/LMS_UT/PEC (leiomyoma / leiomyosarcoma (uterine type) / PEComa) combines tumors of seemingly different morphology and grades, separating the uterine leiomyosarcomas from their counterparts in other localizations. Several sarcomas subsumed under the term “pleomorphic sarcomas” 11 comprising UPS, MFS, pleomorphic liposarcoma and pleomorphic rhabdomyosarcoma are combined in mc UPS/MFS/LS_PL/RMS_PL. RMS_PL (undifferentiated pleomorphic sarcoma / myxofibrosarcoma / pleomorphic liposarcoma / pleomorphic rhabdomyosarcoma) and UPS as well as LS_PL and MFS have been shown to exhibit similar genomic alterations 12 , 13 which is supported by a similar CNV profile of these sarcomas (supplementary Fig. 2). The mc OFD/OFD_ADA (osteofibrous dysplasia / osteofibrous dysplasia like adamantinoma) is clearly separate from mc ADA supporting previous findings on molecular differences between osteofibrous dysplasia-like adamantinoma and adamantinoma 14 . New subclasses within sarcoma entities based on DNA methylation Angiosarcoma (AS) was split up into two methylation subclasses. The subclass AS_A was characterized by a high frequency of MYC amplifications (54%), which were not observed in subclass AS_B (Suppl. Fig.1a). Similar to a previous report, MYC amplification occurred in AS arising in different localizations 15 . Radiation associated AS accumulated in subclass AS_A. No specific clinical feature was associated with AS_B. The number and extent of chromosomal alterations was larger in AS_A. Two separate methylation subclasses for gastrointestinal stromal tumour (GIST) were detected in the reference set, termed GIST_AD (median age 64 years, AD=adult type) and GIST_PED (median age 15 years, PED=pediatric type). The mc GIST_PED likely corresponds to the ( KIT or PDGFRA ) wild-type GIST preferentially seen in young patients supported by the lack of chromosome 14 deletions and a high rate of MGMT gene methylation in this group 16 and a marked female predominance (Suppl. Fig.1b). Among osteosarcoma (OS), a small group of 21 cases termed OS_B separated from the bulk of 189 cases termed OS_A. Copy number variations were clearly lower in OS_B than in OS_A, as well as compared to low grade central osteosarcoma (supplementary Fig. 2). Currently, no clinical outcome data are available for osteosarcoma class OS_B. Patients with histiocytic sarcoma (HS) belonging to mc HS_A and HS_B did not differ in regard to age and tumor localization, however CNV alterations were more numerous in HS_A than in HS_B. Infantile fibrosarcoma (IFS) presented with two distinct methylation classes, IFS_A and IFS_B. Of these, IFS_A formed a cluster together with mesoblastic nephroma (MN) while IFS_B was separate. Both IFS subgroups exhibited comparable CNVs with gains on chromosomal arms 11p/q 15,17p/q and 19p/q being most frequent. Interestingly, gains on chromosome 11 were also shared with MN of the cellular subtype, which show recurrent ETV6-NTRK3 fusions. MPNST included a histologically recognized variant frequently diagnosed as epithelioid MPNST, which exhibits a characteristic methylation profile and typically carries chromosome 2 gains or SMARCB1 homozygous deletions, suggesting the presence of a distinct tumor type provisionally termed epithelioid nerve sheath tumor (ENST). This group shows overlap with a previously reported cohort 17 , although ENST do not exhibit malignancy in all cases. OFMT (ossifying fibromyxoid tumour) displays two different methylation profiles for subclasses OFMT_A and OFMT_B. No distinct clinical characteristics for these methylation subgroups have been defined yet. However, the different CNV profiles for OFMT_A and OFMT_B (supplementary Fig. 2) point towards differences beyond only the methylation profile. Sarcoma subgroups depending on tumor localization Four tumor types, that is angioleiomyoma (ALMO), solitary fibrous tumor (SFT), schwannoma (SWN) and IDH-mutant chondrosarcoma exhibited anatomic location-specific methylation patterns with intracranial tumors clustering separately from peripheral tumors ( Fig. 2b ). Site specific clustering was also observed in leiomyosarcoma, with uterine tumors clustering differently from those in other localizations. In addition to the instances described here, separation of methylation classes based on tumor localization was also previously observed for several brain tumor entities, such as pilocytic astrocytoma and subependymoma. These tumors tended to cluster depending on respective brain compartments 3 . This indicates the importance of tumor location and possible microenvironmental effects and may be of relevance for future classifiers. MDM2 amplification status and DDLS prediction All liposarcomas in the reference set of v13.1 exhibited MDM2 amplifications. We observed classifier prediction of UPS/MFS/LS_PL/RMS_PL in some DDLS cases exhibiting MDM2 amplification. Other tumors with MDM2 amplification were not affected, since their respective methylation profiles were sufficiently different from DDLS. Therefore, we included a caveat in the classification output of v13.1 pointing towards the possibility of DDLS if CNV analysis recognized MDM2 amplification. Specifically, to further support a DDLS classifier prediction, we recommend reviewing the Copy Number Variation (CNV) profile of the tumor for evidence of MDM2 gene amplification. Independent validation of the v13.1 classifier in external sarcoma cohorts Validation of the v13.1 sarcoma classifier was performed employing four independent tumor data sets with a total of 1547 cases. Prior to analyses, all data sets had to pass stringent quality checks. Cases with quality-relevant values below our described thresholds were excluded from the analysis, resulting in a total of 1547 cases passing all quality metrics (see Materials and Methods for details). Samples were defined as within classifier scope (predictable or eligible) if their original histopathological diagnosis aligned with one of the tumor (sub-)classes included in the v13.1 reference cohort. High tumor cell content positively correlated with more cases reaching a classifier prediction (supplementary Fig. 3a). In contrast, the likelihood for a confident classifier prediction to match the original histopathological diagnosis is less dependent on tumor cell content (supplementary Fig. 3b). All cases from the four separate validation data sets are listed in supplementary table 3. A graphic overview, including the distribution of original diagnoses of the validation cohort, is provided in Fig 4a and b . Download figure Open in new tab Figure 4. Independent validation of the v13.1 classifier in external sarcoma cohorts. The classifier was evaluated on 1547 independent samples from four different sources: DKFZ/NCT/DKTK MASTER, INFORM, UCL Biobank, and a prospective cohort from TU-Dresden university hospital. (a) Distribution of cases by source cohort. (b) The 10 most frequent original (institutional) diagnoses across the combined validation cohort. Numbers shown within the donut chart represent the number of samples per diagnosis. (c) Overview of v13.1 sarcoma classifier predictions matching the original diagnoses with different calibrated score thresholds (black bars) across the combined validation cohort, and comparison with the previous sarcoma classifier version (v12.3, grey bars) (d) Distribution of calibrated classifier scores for cases with v13.1 classifier prediction matching original diagnosis, and a red dashed line at 0.9 marking the recommended score threshold. (e) Sankey diagram showing concordance between v13.1 classifier predictions and original diagnoses in the INFORM validation data set, considering the molecular data (NGS-based data for each tumor) for discrepant cases. (f) Sankey diagram showing concordance between v13.1 classifier predictions and original diagnoses in the DKFZ/NCT/DKTK MASTER validation data set, considering the molecular data (NGS- based data for each tumor) for discrepant cases, resulting in classifier-based refinement of original diagnoses for 14 cases in total. Dx = Diagnosis. All 1547 cases from the four validation cohorts were pooled, yielding 1300 cases with tumor types covered by the methylation classes of classifier version v12.3, and 1449 such cases for version v13.1 ( Fig. 4c ). For all cases of the validation data set for which the v13.1 prediction matches the original diagnosis, the calibrated scores for these predictions exceed the recommended threshold of 0.9 in 968/1135 (85%) cases ( Fig. 4c, d ). With the v12.3 classifier, 864 predictions and with the v13.1 classifier, 1061 predictions were reached corresponding to 66% and 73% predictions in cases, for which original histopathological diagnosis aligned with one of the tumor classes included in the v13.1 reference cohort. Concordant diagnosis in eligible cases with predictions (score > 0.9) were obtained in 90,5% employing v12.3 and 91,2% employing the new v13.1 classifier version. INFORM cohort The INFORM cohort enrolls pediatric patients with predominantly recurrent tumors for extensive molecular testing including DNA sequencing, RNA sequencing and methylation analysis 18 . 459 cases of the INFORM cohort with a sarcoma diagnosis were included in the validation set. Of these, 47 cases did not match in their histopathological diagnosis with the tumor types represented in the v13.1 reference set, or the initial diagnosis was non-specific, like “Soft Tissue Sarcoma” for example. INFORM sarcoma samples already included in the reference set were not included in the validation set. Being restricted to pediatric patients, the most frequently represented sarcoma type was Ewing sarcoma followed by osteosarcoma. Data on individual cases were compiled in supplementary table 3. In 311 out of 459 cases (68%), the histopathological diagnosis was supported by a classifier prediction with a score greater than 0.9, thereby increasing diagnostic confidence ( Fig. 4e ). The classifier enabled a definitive molecular diagnosis or tumor reclassification in 25 of 459 (5%) cases with inconclusive or ambiguous histological findings ( Fig. 4e ), providing added value on top of current standard diagnostic practice. MASTER cohort The prospective DKFZ/NCT/DKTK MASTER trial (ClinicalTrials.gov: NCT05852522 ) enrolls adult patients with advanced rare cancers who have exhausted standard treatment options to undergo extensive molecular testing, including whole-genome/exome sequencing, transcriptome sequencing, and DNA methylation profiling, to inform clinical management. 19 . Of the 179 cases in total, 151 were considered eligible (histopathological diagnosis covered in the v13.1 classifier) and among these 75% received a high-score prediction. In 94,6% of the cases, the prediction matched the original diagnosis ( Fig. 4f ). Reevaluation of the clinical diagnosis for 6 non-matching patients based on molecular findings resulted in a change of diagnosis in four patients (marked * in supplementary table 3). Two discrepancies concerning issues in liposarcoma prediction by v13.1 could be resolved based on MDM2 amplification in both cases. Thus, the sarcoma classifier version v13.1 predicted 98,0% of cases correctly in a validation set with strict molecular workup of initially discrepant cases. In the two mismatching cases, v13.1 explicitly hinted towards the problem and suggested consideration of the correct diagnosis of liposarcoma (DDLS). In addition, ten cases for which their original histopathological diagnosis was not included in the v13.1 reference cohort received an updated diagnosis based on molecular data. The updated diagnosis and the v13.1 classifier prediction matched in these 10 cases. In summary, in 106 out of 179 cases (60%), the histopathological diagnosis was supported by a classifier prediction with a score greater than 0.9, thereby increasing diagnostic confidence ( Fig. 4f ). The classifier enabled a definitive molecular diagnosis or tumor reclassification in 14 of 179 (8%) cases with inconclusive or ambiguous histological findings ( Fig. 4f ). UCL Biobank sarcoma set This set is based on publicly accessible data which we have utilized to validate the first version (v12.3) of the sarcoma classifier 20 . The series predominantly contains sarcomas involving bone and cartilage with osteosarcoma and chondrosarcoma being most frequently represented. From the series of 798 cases in total, 775 cases were eligible with a prediction reached in 69%. Prediction and diagnosis matched in 88.9 % of the cases. An overview is given in suppl. table 3. Prospective cohort A series of 111 cases was compiled from prospective new diagnoses after finalizing the v13.1 sarcoma classifier. Only cases with their initial diagnoses represented in the v13.1 sarcoma classifier classes were included. A prediction (with classifier score > 0.9) was reached in 74 % of the cases with 93.9 % of the predictions concordant with the initial diagnosis. An overview is given in supplementary table 3. Discussion Methylation based tumor classification has proven highly valuable in human brain tumors 3 . Non- neuroectodermal tumors, however, may pose a greater challenge for applying this approach than brain tumors. One emerging explanation is the previously underestimated influence of tissue-specific factors on the methylation pattern, as exemplified by the growing number of distinct methylation classes observed for a single tumor entity arising in different organ systems. Typical examples are SFT (solitary fibrous tumor), ALMO (angioleiomyoma) and schwannoma (SWN) with distinct methylation patterns depending on their tumor origin in- or outside the CNS. Two hypotheses may suggest potential origins for the localization specificity observed in methylation classes within tumor entities: firstly, the tumor precursor cells from distinct anatomic locations may exhibit sufficient differences in their methylation pattern to be recognized by the classifier algorithm. Second, site-specific portions of non-tumorous tissues, i.e. the microenvironment, contribute significantly to the methylation pattern. This may require a site-specific approach, with organ-specific classifiers for the vast majority of epithelial tumors, but possibly also for sarcomas, constituting the best solution for future algorithms. In fact, future organ- specific classifiers may even combine epithelial and mesenchymal tumors. Version v12.3 4 of the sarcoma classifier has been subject to multiple independent testing 20 – 22 . While in general considered to be a very promising diagnostic aid, a common concern was the high rate of unclassifiable cases. To this end, v13.1 was expanded to include 93 methylation classes in total, adding 62 new tumor types or subtypes to the previous version. In four validation sets totaling 1547 cases, the rate of predictions increased from 66% in v12.3 to 73% in v13.1. The rate of concordance of prediction and pathological diagnosis increased slightly from 90,5% to 91,2%. The v13.1 classifier comprises more methylation classes that include several tumor types or subtypes at the more fine-grained methylation subclass level than the predecessor, the v12.3 sarcoma classifier. Some entities such as myxofibrosarcoma or desmoplastic melanoma did not separate well from undifferentiated pleomorphic sarcoma or melanoma, respectively, and therefore were not included in v12.3. Inclusion of these entities in v13.1 necessitated the formation of hierarchical methylation classes with contribution from more than a single tumor group. It was thus considered that predicting combined methylation classes based on one or more subordinate subclasses would reduce the risk of misclassification, with further distinction remaining in the expert hands of the pathologist. This approach underlines the concept that the sarcoma classifier is a tool for suggesting diagnostic directions and for aiding decision making, rather than providing a concrete final diagnosis. The new sarcoma classifier was validated using 1527 tumor samples from four independent sarcoma sets of different composition with respect to tumor types. The rate of diagnostic predictions rose to 73% in v13.1 from 66% in v12.3. Of relevance is a strongly increased rate of failed predictions in tumor samples with less than 40% tumor cell content, pointing to the problem of sampling best suited tumor material prior to downstream molecular analysis. The rate of matched predictions increased slightly from 90,5% in v12.3 to 91.2% in v13.1. We applied strict criteria for considering a classifier prediction as consistent with the diagnosis. Without comprehensive molecular data in all cases, however, discrepant predictions cannot be definitively classified as incorrect. In summary, we present a robust, automated classifier leveraging DNA methylation profiling for precise sarcoma diagnosis. The sarcoma classifier v13.1 is accessible at https://app.epignostix.com Supplementary figure 1 a Copy number profiles in two distinct groups of angiosarcoma (AS) and difference in MYC amplification. Myc amplification occurs in mc AS_A while absent in mc AS_B. Chromosome 6q losses and 7 gains are more frequent in mc AS_B. b Copy number profiles in two distinct groups of gastrointestinal tumors and differences in MGMT methylation and sex distribution. Mc GIST_AD exhibits a high frequency of 1p and 14q losses nor seen in mc GIST_PED. Mc GIST_PED occurs predominantly in female patients and tumors exhibit mostly methylation of MGMT. Supplementary figure 2 Summary CNV profiles for 93 methylation classes covered in sarcoma classifier v13.1. X-axis gives chromosomal localization, y-axis provides percentage of tumors affected in the respective methylation class. Supplementary figure 3 Calculated tumor cell content and predictions with a calibrated score >0.9 versus no prediction in 1547 cases of the four validation sets combined. Supplementary table 1 Composition of the v13.1 reference set and comparison with the v12.3 reference set. The table provides abbreviations and full names of the tumor types included in v13.1 (n=4377) and v12.3 (n=1077). Tumors grouped to a single mc are indicated by boxes. * normal non-tumorous control tissues; ** one IFS_B included with IFS in v12.3; *** one LMS_UT included with LMO in v12.3; **** incorrectly termed CHORD_DD in v12.3; ***** removed from Sarcoma reference set because represented in brain tumor classifier 12.5 upward. Supplementary table 2 Description of the methylation classes in v13.1. Amplifications and homozygous deletions observed in 10% or more of the cases within a mc are listed. Only extensive CNV encompassing at least half of the respective chromosomal arm are scored resulting in underscoring of circumscribed alterations which can be estimated in supplementary figure 1. Percentages given for CNV alterations, gene amplifications and homozygous deletions base on analyses of the original idat pairs resulting supplementary tables 3 to 6. All CNV observed in 60% or more of the respective groups are listed. In cases with fewer alterations the most frequently recurring lesion is shown if seen in 30% or more of the cases. Supplementary table 3 1547 cases from four validation sets. Information included is original diagnosis, validation set, predictions of the 12.3 and 13.1 classifiers with calibrated scores and information on whether cases were eligible for the classifiers and whether a prediction was reached and whether the prediction matched with the original diagnosis. Equal Contribution Natalie Jäger and David E. Reuss contributed equally to this work. Support Supported by an Illumina research grant. The MASTER program is supported by the NCT Overarching Clinical Translational Trial Program, the NCT Heidelberg Molecular Precision Oncology Program, and DKTK. Authorś disclosures of potential conflict of interest Natalie Jäger and Martin Sill are full-time employees, Daniel Schrimpf and David TW. Jones are part time employees of Heidelberg Epignostix GmbH. Andreas von Deimling, Felix Sahm, Stefan Pfister, Matija Snuderl, David Capper, David TW. Jones, Daniel Schrimpf and Martin Sill are founders and shareholders of Heidelberg Epignostix GmbH. Stefan Fröhling reports consultancy fees from Illumina. Data sharing statement The training data (n=4377) set has been deposited at GHGA under reference number GHGAS64406795843199 and is available for download (raw idat files) upon signing a data transfer agreement. Data Availability The training data (n=4377) set has been deposited at GHGA under reference number GHGAS64406795843199 and is available for download (raw idat files) upon signing a data transfer agreement. Acknowledgements The authors thank Ulrike Lass, Joshua Trautner, Antje Habel and the DKFZ methylation core facility for skillful technical assistance. Supported by an Illumina research grant. INFORM: We would like to express our sincere thanks to Carsten Maus, Erjia Wang (Genomics and Proteomics Core Facility, DKFZ). Lena Weiser, Gregor Warsow (Omics IT and Data Management Core Facility, DKFZ) for their highly dedicated support in data management and processing. Robert Autry, Gnanaprakash Balasubramanian, Christopher Previti and Rolf Kabbe (Division of Pediatric Neurooncology, DKFZ) for their sincere and dedicated contribution to the bioinformatics analyses. The INFORM program is financially supported by the German Cancer Research Center (DKFZ), several German health insurance companies, the German Cancer Consortium (DKTK), the German Federal Ministry of Education and Research (BMBF), the German Federal Ministry of Health (BMG), the Ministry of Science, Research and the Arts of the State of Baden-Württemberg (MWK BW); the German Cancer Aid (DKH), the German Childhood Cancer Foundation (DKS), RTL television, the aid organization BILD hilft e.V. (Ein Herz für Kinder) and the generous private donation of the Scheu family. We thank the NCT Sample Processing Laboratory and the DKFZ Next Generation Sequencing, Microarray, and Omics IT and Data Management Core Facilities for technical support. References 1. ↵ Ray-Coquard , I. , Montesco , M.C. , Coindre , J.M. , Dei Tos , A.P. , Lurkin , A. , Ranchere-Vince , D. , Vecchiato , A. , Decouvelaere , A.V. , Mathoulin-Pelissier , S. , Albert , S. , et al. ( 2012 ). Sarcoma: concordance between initial diagnosis and centralized expert review in a population-based study within three European regions . Ann Oncol 23 , 2442 – 2449 . doi: 10.1093/annonc/mdr610 . OpenUrl CrossRef PubMed Web of Science 2. ↵ Italiano , A. , Di Mauro , I. , Rapp , J. , Pierron , G. , Auger , N. , Alberti , L. , Chibon , F. , Escande , F. , Voegeli , A.C. , Ghnassia , J.P. , et al. ( 2016 ). Clinical effect of molecular methods in sarcoma diagnosis (GENSARC): a prospective, multicentre, observational study . Lancet Oncol 17 , 532 – 538 . doi: 10.1016/S1470-2045(15)00583-5 . OpenUrl CrossRef PubMed 3. ↵ Capper , D. , Jones , D.T.W. , Sill , M. , Hovestadt , V. , Schrimpf , D. , Sturm , D. , Koelsche , C. , Sahm , F. , Chavez , L. , Reuss , D.E. , et al. ( 2018 ). DNA methylation-based classification of central nervous system tumours . Nature 555 , 469 – 474 , 43,070. doi: 10.1038/nature26000 . OpenUrl CrossRef PubMed 4. ↵ Koelsche , C. , Schrimpf , D. , Stichel , D. , Sill , M. , Sahm , F. , Reuss , D.E. , Blattner , M. , Worst , B. , Heilig , C.E. , Beck , K. , et al. ( 2021 ). Sarcoma classification by DNA methylation profiling . Nat Commun 12 , 498 , 17 .694. doi: 10.1038/s41467-020-20603-4 . OpenUrl CrossRef 5. ↵ Aryee , M.J. , Jaffe , A.E. , Corrada-Bravo , H. , Ladd-Acosta , C. , Feinberg , A.P. , Hansen , K.D. , and Irizarry , R.A . ( 2014 ). Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays . Bioinformatics 30 , 1363 – 1369 . doi: 10.1093/bioinformatics/btu049 . OpenUrl CrossRef PubMed Web of Science 6. ↵ Maros , M.E. , Capper , D. , Jones , D.T.W. , Hovestadt , V. , von Deimling , A. , Pfister , S.M. , Benner , A. , Zucknick , M. , and Sill , M. ( 2020 ). Machine learning workflows to estimate class probabilities for precision cancer diagnostics on DNA methylation microarray data . Nat Protoc 15 , 479 – 512 , 13.491. doi: 10.1038/s41596-019-0251-6 . OpenUrl CrossRef PubMed 7. ↵ Wang , Y. , Chen , S.Y. , Karnezis , A.N. , Colborne , S. , Santos , N.D. , Lang , J.D. , Hendricks , W.P. , Orlando , K.A. , Yap , D. , Kommoss , F. , et al. ( 2017 ). The histone methyltransferase EZH2 is a therapeutic target in small cell carcinoma of the ovary, hypercalcaemic type . J Pathol 242 , 371 – 383 . doi: 10.1002/path.4912 . OpenUrl CrossRef PubMed 8. ↵ Breiman , L . ( 2001 ). Random Forests . Machine Learning 45 , 5 – 32 . OpenUrl CrossRef 9. ↵ Koelsche , C. , Stichel , D. , Griewank , K.G. , Schrimpf , D. , Reuss , D.E. , Bewerunge-Hudler , M. , Vokuhl , C. , Dinjens , W.N.M. , Petersen , I. , Mittelbronn , M. , et al. ( 2019 ). Genome-wide methylation profiling and copy number analysis in atypical fibroxanthomas and pleomorphic dermal sarcomas indicate a similar molecular phenotype . Clin Sarcoma Res 9 , 2 . doi: 10.1186/s13569-019-0113-6 . OpenUrl CrossRef 10. ↵ Bonvalot , S. , Miceli , R. , Berselli , M. , Causeret , S. , Colombo , C. , Mariani , L. , Bouzaiene , H. , Le Pechoux , C. , Casali , P.G. , Le Cesne , A. , et al. ( 2010 ). Aggressive surgery in retroperitoneal soft tissue sarcoma carried out at high-volume centers is safe and is associated with improved local control . Ann Surg Oncol 17 , 1507 – 1514 . doi: 10.1245/s10434-010-1057-5 . OpenUrl CrossRef PubMed Web of Science 11. ↵ Hornick , J.L . ( 2018 ). Subclassification of pleomorphic sarcomas: How and why should we care? Ann Diagn Pathol 37 , 118 – 124 . doi: 10.1016/j.anndiagpath.2018.10.006 . OpenUrl CrossRef PubMed 12. ↵ Barretina , J. , Taylor , B.S. , Banerji , S. , Ramos , A.H. , Lagos-Quintana , M. , Decarolis , P.L. , Shah , K. , Socci , N.D. , Weir , B.A. , Ho , A. , et al. ( 2010 ). Subtype-specific genomic alterations define new targets for soft-tissue sarcoma therapy . Nat Genet 42 , 715 – 721 . doi: 10.1038/ng.619 . OpenUrl CrossRef PubMed 13. ↵ Carvalho , S.D. , Pissaloux , D. , Crombe , A. , Coindre , J.M. , and Le Loarer , F. ( 2019 ). Pleomorphic Sarcomas: The State of the Art . Surg Pathol Clin 12 , 63 – 105 . doi: 10.1016/j.path.2018.10.004 . OpenUrl CrossRef PubMed 14. ↵ Ali , N.M. , Niada , S. , Morris , M.R. , Brini , A.T. , Huen , D. , Sumathi , V. , and Latif , F . ( 2019 ). Comprehensive Molecular Characterization of Adamantinoma and OFD-like Adamantinoma Bone Tumors . Am J Surg Pathol 43 , 965 – 974 . doi: 10.1097/PAS.0000000000001251 . OpenUrl CrossRef PubMed 15. ↵ Hogeboom-Gimeno , A.G. , van Ravensteijn , S.G. , Desar , I.M.E. , Hillebrandt-Roeffen , M.H.S. , van Cleef , P.H.J. , Bonenkamp , J.J. , Flucke , U. , and Versleijen-Jonkers , Y.M.H. ( 2023 ). MYC amplification in angiosarcoma depends on etiological/clinical subgroups - Diagnostic and prognostic value . Ann Diagn Pathol 63 , 152096 . doi: 10.1016/j.anndiagpath.2022.152096 . OpenUrl CrossRef 16. ↵ Giger , O.T. , Ten Hoopen , R. , Shorthouse , D. , Abdullahi , S. , Bulusu , V.R. , Jadhav , S. , Maher , E.R. , and Casey , R.T . ( 2023 ). Preferential MGMT hypermethylation in SDH-deficient wild-type GIST . J Clin Pathol 77 , 34 – 39 . doi: 10.1136/jcp-2022-208462 . OpenUrl CrossRef PubMed 17. ↵ Schaefer , I.M. , Dong , F. , Garcia , E.P. , Fletcher , C.D.M. , and Jo , V.Y . ( 2019 ). Recurrent SMARCB1 Inactivation in Epithelioid Malignant Peripheral Nerve Sheath Tumors . Am J Surg Pathol 43 , 835 – 843 . doi: 10.1097/PAS.0000000000001242 . OpenUrl CrossRef PubMed 18. ↵ van Tilburg , C.M. , Pfaff , E. , Pajtler , K.W. , Langenberg , K.P.S. , Fiesel , P. , Jones , B.C. , Balasubramanian , G.P. , Stark , S. , Johann , P.D. , Blattner-Johnson , M. , et al. ( 2021 ). The Pediatric Precision Oncology INFORM Registry: Clinical Outcome and Benefit for Patients with Very High-Evidence Targets . Cancer Discov 11 , 2764 – 2779 , 38.272. doi: 10.1158/2159-8290.CD-21-0094 . OpenUrl Abstract / FREE Full Text 19. ↵ Jahn , A. , Rump , A. , Widmann , T.J. , Heining , C. , Horak , P. , Hutter , B. , Paramasivam , N. , Uhrig , S. , Gieldon , L. , Drukewitz , S. , et al. ( 2022 ). Comprehensive cancer predisposition testing within the prospective MASTER trial identifies hereditary cancer patients and supports treatment decisions for rare cancers . Ann Oncol 33 , 1186 – 1199 . doi: 10.1016/j.annonc.2022.07.008 . OpenUrl CrossRef PubMed 20. ↵ Lyskjaer , I. , De Noon , S. , Tirabosco , R. , Rocha , A.M. , Lindsay , D. , Amary , F. , Ye , H. , Schrimpf , D. , Stichel , D. , Sill , M. , et al. ( 2021 ). DNA methylation-based profiling of bone and soft tissue tumours: a validation study of the ’DKFZ Sarcoma Classifier’ . J Pathol Clin Res 7 , 350 – 360 , 4.373. doi: 10.1002/cjp2.215 . OpenUrl CrossRef PubMed 21. Miettinen , M. , Abdullaev , Z. , Turakulov , R. , Quezado , M. , Luina Contreras , A. , Curcio , C.A. , Rys , J. , Chlopek , M. , Lasota , J. , and Aldape , K.D. ( 2024 ). Assessment of The Utility of The Sarcoma DNA Methylation Classifier In Surgical Pathology . Am J Surg Pathol 48 , 112 – 122 . doi: 10.1097/PAS.0000000000002138 . OpenUrl CrossRef PubMed 22. ↵ Roohani , S. , Ehret , F. , Perez , E. , Capper , D. , Jarosch , A. , Florcken , A. , Mardian , S. , Zips , D. , and Kaul , D . ( 2022 ). Sarcoma classification by DNA methylation profiling in clinical everyday life: the Charite experience . Clin Epigenetics 14 , 149 . doi: 10.1186/s13148-022-01365-w . OpenUrl CrossRef View the discussion thread. Back to top Previous Next Posted June 30, 2025. Download PDF Supplementary Material Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Advancing sarcoma diagnostics with expanded DNA methylation-based classification Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Advancing sarcoma diagnostics with expanded DNA methylation-based classification Natalie Jäger , David E. Reuss , Martin Sill , Daniel Schrimpf , Abigail K. Suwala , Philipp Sievers , Rouzbeh Banan , Felix Hinz , Ramin Rahmanzade , Henry Bogumil , Kaan Fuat Aras , Areeba Patel , Andrey Korshunov , Melanie Bewerunge-Hudler , Arjen HG. Cleven , Manel Esteller , Hanno Glimm , Wolfgang Hartmann , Simon Kreutzfeld , Christoph Heilig , Till Milde , Iver Petersen , Christian M. Vokuhl , Wolfgang Wick , Olaf Witt , Thibault Kervarrec , Evelina Miele , Jonathan Serrano , Stephan Frank , Karl Kashofer , Anne Mc Leer , Elke Pfaff , Melanie Pages , Arnault Tauziede-Espariat , Ferdinand Toberer , Henning B. Boldt , Petr Martinek , Sebastian Brandner , Mayara Euzebio , Aurore Siegfried , Jane Chalker , Patrik Harter , Romain Appay , Wolfgang Dietmaier , Martin Hasselblatt , Uta E. Flucke , Laura S. Hiemcke-Jiwa , David Solomon , Clara Frydrychowicz , Pascale Varlet , Benjamin Goeppert , Michaela Nathrath , Claudia Blattmann , Monika Sparber-Sauer , August Kolb , Michel Mittelbronn , Thomas Mentzel , Sandra Leisz , Anja Harder , Till Acker , Drew Pratt , Eva Wardelmann , Jamal Benhamida , Mark Ladanyi , Philipp Jurmeister , William Foulkes , Pamela Ajuyah , David Z. Ziegler , Jürgen Hench , Maikel JL. Nederkoorn , Yvonne MH. Versleijen-Jonkers , Gunhild Mechtersheimer , Sandro Krieg , Manfred Gessler , Daniel Baumhoer , Sam Behjati , Luca Bertero , Klaus Griwank , Dirk Schadendorf , Pancras CW. Hogendoorn , Jean-Francois Emile , Paul G. Kemps , Armin Jarosch , Michael W. Ronellenfitsch , Toni Su Idler , Daniela Aust , Sylvia Herold , Jessica Pablik , Maysa Al-Hussaini , Zied Abdullaev , Maximus Yeung , Marco Wachtel , Eva Brack , Felix KF. Kommoss , Markku Miettinen , Ken Aldape , Adrienne MH. Flanagan , Uta Dirksen , Kristian Pajtler , Thomas GP. Grünewald , Daniel Lipka , Stefan Fröhling , Christian Koelsche , Matija Snuderl , David Capper , Stefan M. Pfister , David TW. Jones , Felix Sahm , Andreas von Deimling medRxiv 2025.06.30.25330543; doi: https://doi.org/10.1101/2025.06.30.25330543 Share This Article: Copy Citation Tools Advancing sarcoma diagnostics with expanded DNA methylation-based classification Natalie Jäger , David E. Reuss , Martin Sill , Daniel Schrimpf , Abigail K. Suwala , Philipp Sievers , Rouzbeh Banan , Felix Hinz , Ramin Rahmanzade , Henry Bogumil , Kaan Fuat Aras , Areeba Patel , Andrey Korshunov , Melanie Bewerunge-Hudler , Arjen HG. Cleven , Manel Esteller , Hanno Glimm , Wolfgang Hartmann , Simon Kreutzfeld , Christoph Heilig , Till Milde , Iver Petersen , Christian M. Vokuhl , Wolfgang Wick , Olaf Witt , Thibault Kervarrec , Evelina Miele , Jonathan Serrano , Stephan Frank , Karl Kashofer , Anne Mc Leer , Elke Pfaff , Melanie Pages , Arnault Tauziede-Espariat , Ferdinand Toberer , Henning B. Boldt , Petr Martinek , Sebastian Brandner , Mayara Euzebio , Aurore Siegfried , Jane Chalker , Patrik Harter , Romain Appay , Wolfgang Dietmaier , Martin Hasselblatt , Uta E. Flucke , Laura S. Hiemcke-Jiwa , David Solomon , Clara Frydrychowicz , Pascale Varlet , Benjamin Goeppert , Michaela Nathrath , Claudia Blattmann , Monika Sparber-Sauer , August Kolb , Michel Mittelbronn , Thomas Mentzel , Sandra Leisz , Anja Harder , Till Acker , Drew Pratt , Eva Wardelmann , Jamal Benhamida , Mark Ladanyi , Philipp Jurmeister , William Foulkes , Pamela Ajuyah , David Z. Ziegler , Jürgen Hench , Maikel JL. Nederkoorn , Yvonne MH. Versleijen-Jonkers , Gunhild Mechtersheimer , Sandro Krieg , Manfred Gessler , Daniel Baumhoer , Sam Behjati , Luca Bertero , Klaus Griwank , Dirk Schadendorf , Pancras CW. Hogendoorn , Jean-Francois Emile , Paul G. Kemps , Armin Jarosch , Michael W. Ronellenfitsch , Toni Su Idler , Daniela Aust , Sylvia Herold , Jessica Pablik , Maysa Al-Hussaini , Zied Abdullaev , Maximus Yeung , Marco Wachtel , Eva Brack , Felix KF. Kommoss , Markku Miettinen , Ken Aldape , Adrienne MH. Flanagan , Uta Dirksen , Kristian Pajtler , Thomas GP. Grünewald , Daniel Lipka , Stefan Fröhling , Christian Koelsche , Matija Snuderl , David Capper , Stefan M. Pfister , David TW. Jones , Felix Sahm , Andreas von Deimling medRxiv 2025.06.30.25330543; doi: https://doi.org/10.1101/2025.06.30.25330543 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Oncology Subject Areas All Articles Addiction Medicine (568) Allergy and Immunology (863) Anesthesia (299) Cardiovascular Medicine (4425) Dentistry and Oral Medicine (443) Dermatology (382) Emergency Medicine (607) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1507) Epidemiology (15221) Forensic Medicine (30) Gastroenterology (1123) Genetic and Genomic Medicine (6588) Geriatric Medicine (667) Health Economics (997) Health Informatics (4524) Health Policy (1368) Health Systems and Quality Improvement (1612) Hematology (540) HIV/AIDS (1264) Infectious Diseases (except HIV/AIDS) (15910) Intensive Care and Critical Care Medicine (1103) Medical Education (623) Medical Ethics (145) Nephrology (667) Neurology (6588) Nursing (346) Nutrition (998) Obstetrics and Gynecology (1143) Occupational and Environmental Health (956) Oncology (3331) Ophthalmology (970) Orthopedics (369) Otolaryngology (420) Pain Medicine (435) Palliative Medicine (129) Pathology (663) Pediatrics (1690) Pharmacology and Therapeutics (691) Primary Care Research (710) Psychiatry and Clinical Psychology (5440) Public and Global Health (9219) Radiology and Imaging (2195) Rehabilitation Medicine and Physical Therapy (1369) Respiratory Medicine (1196) Rheumatology (593) Sexual and Reproductive Health (710) Sports Medicine (529) Surgery (710) Toxicology (99) Transplantation (289) Urology (265) (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ffa7f30a8a506cf',t:'MTc3OTQzOTAyNA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();