Dorothée Selimoglu-Buet, Sylvie Chevret, Valeria Santini, Sylvain Thépot, Margot Morabito, Lionel Adès, Lucie Laplane, Aristoteles Giagounidis, Nathalie Droin, Michael Lübbert, Rosa Sapena, Stanislas Nimubona, Jose Miguel Torregrosa‐Diaz, Ulrich Germing, Anna Maria Pelizzari, Sophie Park, Nadja Jaekel, Georgia Metzgeroth, Francesco Onida, Franciane Paul, Andrea Patriarca, Aspasia Stamatoullas, Katharina S. Götze, Fatiha Chermat, Uwe Platzbecker, Pierre Fenaux, Éric Solary, Raphaël Itzykson
Whether mitigation of myeloproliferation improves prognosis of CMML independently of bone marrow response is unknown. Flow-defined classical monocytes (cMo) and immature granulocytes (iGRAN) have not yet been studied as biomarkers of response. We inspected the prognostic value of WBC, circulating monocytes, cMo and iGRANs in the 120 DACOTA (NCT02214407) patients randomized to decitabine ( n = 63) or hydroxyurea ( n = 57) evaluated after 3 cycles with BM aspiration and complete blood count. Across arms, 59% and 56% patients had monocytes > 1 × 10 9 /L or WBC > 10 × 10 9 /L at the 3- and 6-cycle evaluation respectively. After 6 cycles, persistence of monocytes > 1 × 10 9 /L or WBC > 10 × 10 9 /L increased the hazard of death (HR = 5.38, p = 0.0003) irrespective of treatment, baseline CPSS and persistence of BM blast excess. After 3 cycles, both higher absolute cMo and iGRAN counts independently predicted poorer OS, without significant interaction with treatment arm. Median OS from landmark was 35.1 months in the 28% patients with cMo ≤ 0.94 ×10 9 /L AND iGRAN ≤ 0.40 ×10 9 /L versus 15.3 months in others ( p = 0.013). Biomarkers integrating blood counts and flow cytometry may predict CMML prognosis irrespective of treatment.