Pb2027: Prognostication in the Everyday Myeloma Clinic. A Systematic Review of Prognostic Models Validated in Real World Populations

Abstract

Background:Background: Multiple Myeloma (MM) is a severe malignancy with changing treatment possibilities and generally improved survival. However, prognosis is variable, and solid prognostic models for use in the daily clinic are warranted. Several new prognostic models and frailty scores have been introduced, but it can be difficult to keep track of their performance and validation. Aims:Aims: To help clinicians prognosticate patients with newly diagnosed MM in the daily clinic, we performed a systematic review of studies validating prognostic models in general MM populations. Methods:Methods: A review was performed according to the Preferred Reporting Items for Systematic Review and Meta-Analysis guidelines (PRISMA-P) and pre-registered in the International Prospective Register of Systematic Review (PROSPERO) -identification number: CRD42021230604 https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=230604 . Search strategyLouise Redder, Tine Rosenberg, and Niels Abildgaard created the search strategy in collaboration with a Health Sciences Librarian from the University of Southern Denmark. A broad search string was set up with four search blocks: the first included a validated search filter for finding prognostic and diagnostic predictions studies. The second block searched all MeSH terms for MM. The third block contained an explorative search for progression free survival (PFS) and overall survival (OS), as these were our comparative outcomes and an inclusion criterion. In the fourth block, we attempted to target MM populations as close to the real world as possible, through an explorative search for population, cohort studies, multicenter or single center, and registry.Two systematic literature searches were conducted on the 18 th of September 2020 and the 15 th of December 2021, using the electronic bibliographic databases, MEDLINE, EMBASE, and the Cochrane Library. Inclusion & exclusion criteriaStudies validating or replicating prognostic models predicting OS and/or PFS in a population-based cohort of patients with newly diagnosed MM were included. Cohorts composed by fulfilling inclusion and exclusion criteria in clinical trials were excluded. Further, we excluded cohorts based on specific ethnicities, socioeconomic status, somatic diseases or disabilities, and populations representing other plasma cell dyscrasia, such as plasma cell leukaemia or smouldering MM.A prognostic model was defined as a combination of at least two separate variables predicting patient outcome. The model had to predict OS/PFS at the time of diagnosis.The design could be cohort, nested case-control or case-cohort studies as they are recommended for prognostic model