Study type

Study topic

Human medicinal product
DiseaseĀ /health condition

Study type

Non-interventional study

Scope of the study

Assessment of risk minimisation measure implementation or effectiveness

Data collection methods

Secondary use of data
Non-interventional study

Non-interventional study design

Other

Non-interventional study design, other

Retrospective analysis of randomized controlled trial data using machine learning
Study drug and medical condition

Study drug International non-proprietary name (INN) or common name

CARFILZOMIB

Medical condition to be studied

Cardiac failure
Hypertension
Arrhythmia
Pulmonary hypertension
Population studied

Short description of the study population

The study population consists of both (Kyprolis treatment) arms of the ARROW study (Moreau 2018) and the Kyprolis treatment arms of the ASPIRE (Stewart 2015), ENDEAVOR (Dimopoulos 2016) and FOCUS studies (Hajek 2017).
Inclusion Criteria
The subjects in this study are the individuals who participated in the ARROW, ASPIRE, ENDEAVOR and FOCUS studies and were treated with carfilzomib.

Age groups

Adults (18 to < 46 years)
Adults (46 to < 65 years)
Adults (65 to < 75 years)
Adults (75 to < 85 years)
Adults (85 years and over)

Special population of interest

Other

Special population of interest, other

Patients with Cardiac failure, Hypertension, Arrhythmia, Pulmonary hypertension

Estimated number of subjects

1485
Study design details

Main study objective

Develop and characterize risk profiles for select cardiovascular adverse events in patients with relapsed and/or refractory (R/R) multiple myeloma (MM) treated with carfilzomib across four clinical trials through an analysis of baseline characteristics.

Outcomes

Rate of CVAEs in cohorts of patients identified as having high, intermediate or low risks for the select CVAEs

Data analysis plan

The analysis uses three techniques in sequence:Topological data analysis to produce network representations of data,Network clustering known as cold-spot detection to identify coherent sets of non-AE subjects, andMulti-class single-decision-tree learning to discover groups of subjects and conditions on variables that explain them.The sequence may be repeated more than once.
Documents
Study results
English (1.19 MB - PDF)View document