Biomedical research institute
    You are here: Home / PhD program / Thesis defense / Targeted molecular markers for the prognosis of adrenocortical carcinoma: from genomics to personalized medicine

    Targeted molecular markers for the prognosis of adrenocortical carcinoma: from genomics to personalized medicine

    October 13th 16:00 pm 2020

    Rosalind Franklin room, 2nd floor

    Institut Cochin, 22 rue Méchain, Paris 75014


    Supervisor; Guillaume Assié

    Team: Genomics and Signaling of Endocrine Tumors

    Department of Endocrinology, Metabolism and Diabetes (EMD)


    The prognosis of adrenocortical carcinoma (ACC) is globally poor, with 5-year overall survival below 40%, but varies widely. Prognostic stratification is critical for patients’ care. Standard prognostic factors mainly include tumor stage and Ki67 proliferation index. However, the risk stratification based on tumor these factors is limited.

    Recently, pangenomic studies have idenfied ACC subgroups characterized by distinct molecular alterations and associated with different outcomes. A first ACC subgroup is characterized by “C1A” transcriptome profile (proliferation signature), “CIMP” (CpG islands methylator phenotype) methylome profile, “Noisy” chromosome alterations profile (high number of chromosome alterations) and recurrent mutations in p53/Rb and Wnt/β-catenin genes. This subgroup is associated with poor prognosis. Conversely, another ACC subgroup is characterized by “C1B” transcriptome profile (immune signature), “non-CIMP” methylome profile, “Chromosomal” (extended loss of heterozygosity) or “Quiet” (limited number of alterations) chromosome alterations profile and low mutational burden. This subgroup is associated with better prognosis. A targeted marker derived from transcriptome - BUB1B-PINK1 expression measured by RT-qPCR – was proposed for ACC prognostication.


    In line with these previous studies, this thesis aims at developing targeted molecular markers for routine prognostic assessment.

    First, 4-gene methylation (PAX5, PAX6, PYCARD, GSTP1) measured by MS-MLPA was proposed as a targeted prognostic marker derived from methylome.

    Then, the molecular classification was summarized using two combinations of targeted molecular markers: 1) a three dimensional (3D)-targeted classifier, based on tumor RNA and DNA, combining BUB1B-PINK1 expression, PAX5-GSTP1-PYCARD-PAX6 methylation, and targeted measurement of chromosome alterations ; 2) a DNA-based targeted classifier, using tumor DNA only, combining  PAX5-GSTP1-PYCARD-PAX6 methylation, targeted measurement of chromosome alterations, and somatic mutations of p53/Rb and Wnt/β-catenin genes. In localized ACC, combination of tumor stage, tumor-proliferation index, and molecular class provided the most discriminant prognostic model.

    In a third part, we investigated intratumor heterogeneity of DNA-based prognostic markers. Somatic mutations, and, in a less extent, targeted methylation measurement, could vary from one tumor region to another. Therefore, combination of multiple targeted molecular markers, along with clinical features, should be preferred to gene alterations profile alone for the prognostic assessment of ACC.

    In the last part, we aimed to determine transcriptome profiles from FFPE samples, using a dedicated protocol of 3’ RNA-sequencing. 3’ ends are indeed more resistant to paraffin-induced RNA degradation. The 3’ RNA-sequencing protocol successfully classified “C1A” and “C1B” transcriptome profiles, and represents a convenient solution for transcriptome-based prognostic classification in FFPE samples.


    In conclusion, molecular class is established as an essential prognostic factor in localized ACC. Several targeted markers, accounting for the differences in material and techniques availibilty between centers, are proposed for the routine determination of molecular classification. Finally, the determination of transcriptome in FFPE samples paves the way to the integration of genomic medicine into routine care.


    Keywords : adrenocortical carcinoma, prognosis, genomics, transcriptome, methylome.