Targeted Molecular Markers Derived from Genomic Classification for Adrenocortical Cancer  Prognostication

Presentation Number: OR01-4
Date of Presentation: April 3rd, 2017

Guillaume Assie*1, Anne Jouinot2, Martin Fassnacht3, Rossella Libe1, Bertrand Dousset4, Silviu Sbiera3, Cristina Lucia Ronchi3, Matthias Kroiss3, Esther Korpershoek5, Ronald R de Krijger6, Jens Waldmann7, Marcus Quinkler8, Antoine Tabarin9, Olivier Chabre10, Michaela Luconi11, Massimo Mannelli12, Lionel Groussin13, Eric Baudin14, Laurence Amar15, Felix Beuschlein16 and Jerome Yves Bertherat17
1INSERM U 1016, CNRS 8104, Institut Cochin, Paris Descartes University, Paris, France, 2INSERM U1016, CNRS UMR 8104, Institut Cochin Université Paris Descartes, Paris, 3University Hospital Wuerzburg, Wuerzburg, Germany, 4Department of Digestive and Endocrine Surgery, Assistance Publique Hôpitaux de Paris, Hôpital Cochin, Paris, France, 5Dept. of Pathology, Erasmus MC University Medical Center, Rotterdam, The Netherlands, 6Erasmus MC, Rotterdam, Netherlands, 7Visceral-, Thoracic and Vascular Surgery, University Hospital Giessen and Marburg, Marburg, Germany, 8Charité University Medicine Berlin, Campus Mitte, Berlin, Germany, 9University Hospital (CHU) of Bordeaux, Pessac, France, 10Grenoble University Hospital, Grenoble, France, 11Department of Experimental and Clinical Biomedical Sciences, University of Florence Florence, Italy, Florence, Italy, 12Univ of Florence, Florence, Italy, 13INSERM U 1016, CNRS 8104, Institut Cochin, Paris Descartes University, Paris, 14Institut Gustave-Roussy, Villejuif, France, 15Hopital Europeen Georges Pompido, Paris, France, 16Klinikum der Universität München, Ludwig-Maximilian University, Munich, Germany, 17INSERM U 1016, Cochin Institute, Paris Descartes University, Paris, France

Abstract

Background: Adrenocortical cancer (ACC) is an aggressive tumour with heterogeneous prognosis. Recently integrated genomics reported distinct genomic alterations. Transcriptome discriminates “C1A” -with high expression of proliferation / cell cycle-related genes- vs “C1B” ACCs. Methylome discriminates “CIMP” -with hypermethylation in CpG islands- vs “non CIMP” ACCs. Chromosome alterations discriminate “noisy”–with numerous and anarchic alterations-, vs “chromosomal”–with extended patterns of loss of heterozygosity- and “quiet” ACCs –with limited alterations-. Exome identified recurrent mutations in about 20 genes in 60% of ACCs.

These genomic alterations converge into a single 3-groups classification, correlating with outcome. Poor, intermediate and better outcome are associated with “C1A” / “CIMP” / “noisy” ACCs, “C1A”/ “non CIMP”, and “C1B” / “non CIMP” respectively.

The aim was to develop and validate targeted molecular markers reflecting these genomic groups.

Patients and methods: 245 ACC were included from 21 ENSAT (European network for the Study of Adrenal Tumors) centers. Tumor RNA was assessed by RT-qPCR for BUB1B-PINK1 expression (n=135; TaqMan assay, ThermoFisher). Tumor DNA was studied by SNP array for chromosomal alterations (n=238; HumanCore, Illumina), targeted NGS for mutations of 20 driver genes (n=251; AmpliSeq and PGM, Thermofisher), and MS-MLPA for CpG islands methylation of 4 genes -PAX5, GSTP1, PYCARD, PAX6- (n=221; Salsa MLPA EK1, MRC-Holland).

Results:

Gene expression levels identified “C1A”, “C1B” and “undetermined” ACCs in 50, 38 and 12% of cases respectively.

Methylome identified “CIMP” and “non CIMP” ACCs in 44 and 56% of cases respectively.

Chromosomal alteration profiles identified “chromosomal”, “noisy” and “quiet” ACCs in 55, 32 and 13% of cases respectively.

Recurrent mutations or deletions were found in ZNFR3, CDKN2A, TP53 and CTNNB1 in 21, 17, 15 and 11% of cases respectively, in agreement with previous exome studies.

All molecular statuses were available for 107 tumors. 87/107 (81%) concordantly recapitulated the 3 main ACC subgroups previously identified by genomic classifications: 37 were “C1A” / “CIMP”, 18 were “C1A” / “non CIMP”, and 32 were “C1B” / “non CIMP”. Death events occurred in 34/37, 7/18 and 1/32 patients respectively (Fisher exact p<10-15).

Conclusion: Targeted molecular measures can recapitulate the genomic classification of ACC, giving original and useful prognostication information for patient management.

 

Disclosure: MQ: , Shire, , Shire. Nothing to Disclose: GA, AJ, MF, RL, BD, SS, CLR, MK, EK, RRD, JW, AT, OC, ML, MM, LG, EB, LA, FB, JYB