SCIENTIFIC PROGRAMME
SESSION I
HOW I TREAT
SMOLDERING MYELOMA
(SMM)
SESSION II
HOW I TREAT NEWLY
DIAGNOSED MULTIPLE
MYELOMA
SESSION III
FROM RISK
STRATIFICATION TO
RISK-BASED THERAPY?
DEBATE 1
SHOULD WE USE MRD
TESTING TO DETERMINE
THERAPY IN MULTIPLE
MYELOMA?
DEBATE 2
IS THERE A FUTURE ROLE
OF AUTOLOGOUS STEM
CELL TRANSPLANTATION?
SESSION IV
HOW I TREAT RELAPSED
MULTIPLE MYELOMA
DEBATE 3
SHOULD EVERY PATIENT
RECEIVE DARATUMUMAB
IN FIRST LINE?
ROUNDTABLE
MULTIPLE MYELOMA
FROM THE PERSPECTIVE
OF FDA/EMEA AND
FOUNDATIONS
SESSION V
YOU CAN’T BE IMMUNE
FOR IMMUNE THERAPY
ANYMORE
SESSION VI
OTHER PLASMA CELL
DYSCRASIAS
KEYNOTE LECTURES
THE FUTURE OF
MULTIPLE MYELOMA
SELECTED ABSTRACTS
FOR AN ORAL
PRESENTATION
ABSTRACTS SELECTED
AS POSTERS
DISCLOSURES POSTER 1
IMPROVING THE RISK STRATIFICATION OF MULTIPLE MYELOMA WITH A DNA/RNA
EDITORS-BASED SCORING SYSTEM
Sarah Grasedieck1, Afsaneh M. Shariatpanahi2, Matthew C. Jarvis3, Faezeh Borzooee4, Reuben S.
Harris3, Mani Larijani4, Kevin Song5, Arefeh Rouhi2 and Florian Kuchenbauer2
(1)Canada's Michael Smith Genome Sciences Centre, British Columbia Cancer, Vancouver, Canada
(2)Terry Fox Laboratory, British Columbia Cancer, Vancouver, Canada
(3)Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota,
Minneapolis
(4)Department of Molecular Biology and Biochemistry, Faculty of Science, Simon Fraser University,
Burnaby, Canada
(5)Leukemia/Bone Marrow Transplant Program of British Columbia, Vancouver General Hospital, BC
Cancer, Vancouver, Canada
Objective: Clinical prognostication of newly diagnosed MM (NDMM) patients solely relies on clinical
parameters and/or recurrent genetic changes, which only reflect early events in the development of
MM. However, recent insights into the pathogenesis of MM highlighted genome editing through
APOBEC/ADAR genes as well as inflammation as drivers for the onset and progression of MM.
Method: Based on these findings, we hypothesized that a prognostic score reflecting biological
processes as well as clinical features is superior to all currently used classification systems for MM
patients, such as ISS, R-ISS and the Mayo clinic classification. The Multiple Myeloma Research
Foundation CoMMpass study genomics dataset, combining WGS, mRNA Seq and clinical data of more
than 700 patients, allowed us to evaluate prognostic values of demographic and clinical parameters,
cytogenetics, and gene expression levels of ADAR and the APOBEC family as well as inflammation
modulating cytokines of MM patients. We calculated univariate Cox proportional hazards ratios and
Kaplan-Meier survival estimates for all extracted features. In combining clinical variables that were
associated with PFS and OS, we then used multivariate analysis and machine learning approaches to
identify the most accurate PFS and OS model based on C-index and AUROC to define a new risk score
that better stratify NDMM patients.
Result: Machine learning approaches defined a weighted OS/PFS risk score (AI score) consisting of
APOBEC2, APOBEC3A , APOBEC3B, APOBEC3C, ADAR1, IL10, TGFB1, ß-2-microglobulin, hemoglobin,
LDH and creatinine, that achieved the best classification outcome and showed superior performance
compared to commonly used risk classifiers such as ISS and R-ISS. Of note, the applied machine learning
algorithms excluded recurrent genetic changes, such as translocations, based on their relative low
weight from the AI score. Besides a superior overall risk stratification, the AI score further allowed to
identify subgroups of MM patients with very good prognosis (5-year OS: 97%) and an ultra- high risk
group of patients (2-year OS: 59%) who do not benefit from maintenance therapy.
Conclusion: Our findings support the adoption of molecular biomarkers, reflecting dynamic biological
processes rather than cytogenetics for a more accurate risk classification of MM. Considering that
mRNA Seq is as cost effective as FISH and is being increasingly adopted by hospitals, the AI score is a
unique approach to shed light on underlying molecular mechanisms that drive disease progression and
the development of true risk-adapted treatment strategies.