Predictive Model of Early Relapse in Newly Diagnosed Multiple Myeloma: Analysis from a Pooled Dataset

• Published in: Blood 2019-11, Vol.134 (Supplement_1), p.2130-2130
• Main Authors: Zaccaria, Gian Maria, Capra, Andrea, Petrucci, Maria Teresa, Offidani, Massimo, Montefusco, Vittorio, Di Raimondo, Francesco, Bernardini, Annalisa, Liberati, Anna Marina, Omedé, Paola, Mannina, Donato, Caravita di Toritto, Tommaso, Muccio, Vittorio Emanuele, Cascavilla, Nicola, Montani, Stefania, Patriarca, Francesca, Benevolo, Giulia, Saraci, Elona, Belotti, Angelo, Gaidano, Gianluca, Specchia, Giorgina, Nagler, Arnon, Hajek, Roman, Spencer, Andrew, Sonneveld, Pieter, Boccadoro, Mario, Gay, Francesca
• Format: Article
• Language: English
• ISSN: 0006-4971; 1528-0020
• DOI: 10.1182/blood-2019-125640

Background. Despite the improvement of therapeutic regimens, a relevant proportion of multiple myeloma (MM) patients (pts) experience early relapse (ER) [Majithia, Leukemia 2016] and represents an unmet medical need. It is therefore of high clinical interest to identify baseline factors that may predict ER. AimsTo design models predictive of ER (defined as pts with a time-to-progression ≤18 or ≤ 24 months).To assess the accuracy of every model on an independent validation set.To build a score to predict ER. Methods Data were obtained from 2326 pts enrolled in 8 multi-center clinical trials: NCT01093196, NCT01346787, NCT01857115, NCT01190787, NCT00551928, NCT01091831, NCT01063179 and 2005-004730-41. Here, we included 14 baseline features (fts): age, creatinine, albumin (alb), b2microglobulin (b2m), bone marrow plasma cell (PCbm) were evaluated as continuous variables; free light chain (FLC, λ vs K), M-component subtype (IgA vs others), Revised International Staging System (R-ISS stage II/III vs I), lactate dehydrogenase levels >/≤ upper limit of normal (LDHULN), presence vs absence of chromosomal abnormalities detected by FISH [del17p, t(4;14), t(14;16), t(11;14)], and presence of plasmacytomas as categorical values. Trials were assigned to training and validation set to have a superimposed median (μe) age and follow-up in the two subsets. From the training set, a univariate analysis (UV) on outcome was performed according to both Chi-square and Kruskal tests, as appropriate. Features with p