The Proinflammatory Protein S100A9 Suppresses Erythropoietin Elaboration in Patients with Myelodysplastic Syndromes

Introduction: Accumulating evidence implicates innate immune activation in the pathobiology of myelodysplastic syndromes (MDS) and its inflammatory bone marrow microenvironment. Excess elaboration of S100A9 accompanied by secondary induction of TNFα and IL-1β, characterize the inflammatory milieu found in lower risk MDS. Erythroid stimulating agents (ESA) and lenalidomide (LEN) are erythropoietic promoters with known anti-inflammatory properties. To date, the role of such inflammation parameters in endogenous erythropoietin (Epo) regulation and response to such treatments has not been investigated. Herein, we investigated the role of these inflammatory cytokines on Epo elaboration and response to ESA and LEN in MDS.

Materials and Methods: The HepG2 hepatoma cell line was used to investigate Epo elaboration in vitro. Serum collected from 311 MDS patients from 3 centers (AOU Careggi, University of Firenze; Saint Louis hospital in Paris; H. Lee Moffitt Cancer Center in Tampa) were investigated. 125 were obtained prior to ESA treatment and 186 prior to LEN or LEN+EPO treatment. ELISAs for S100A9, S100A8, TNFα, IL1β and EPO were performed. Clinical data were analyzed from all patients. Spearman's correlation, Mann-Withney and Jonckheere-Terpstra tests were used for analysis of continuous variables. Chi-square test was used for analysis of non-continuous variables.

Results: Median age of the patients was 76 years [41-94]. IPSS was low, intermediate-1, intermediate-2 and high risk in 47%, 49%, 3% and 1% of patients, respectively; whereas 28%, 42%, 22%, 7% and 1% of patients were very low, low, intermediate, high and very high risk according to IPSS-R. Serum S100A9 concentration was significantly higher in patients with lower risk versus higher risk MDS (17,798 pg/ml vs 15,291 pg/ml, p=0.01). No significant difference was observed for TNFα and IL1 β according to IPSS or IPSS-R. ELISA quantitation following 24h in vitro stimulation of HepG2 cells with S100A9 (10 to 20 µg/ml), TNFα (10 to 100 ng/ml) or IL1β (10 to 100ng/ml) showed that each cytokine significantly suppressed Epo elaboration in a concentration-dependent manner. To validate these findings, we assessed the relationship between serum concentration of these inflammatory cytokines and Epo level in MDS patients. We observed a significant negative correlation between TNFα and Epo concentration (r=-0.158, p=0.01), and a corresponding significant negative correlation between S100A9 and Epo (r=-0.143, p=0.01). We also found a significant positive correlation between S100A9 and S100A8 (r=0.757, p<0.001), S100A9 and TNFα (r=0.222, p<0001), S100A9 and IL1β (r=175, p=0.004), and IL1β with TNFα concentration (r=257, p<0.001). Moreover, we found that ESA responders had a significant higher serum TNFα concentration prior to treatment than non-responders (12 pg/ml vs 6 pg/ml, p=0.02) with a corresponding significantly lower Epo concentration in ESA responders. There was no significant correlation between S100A9 serum concentration and EPO response. Finally, in patients treated with LEN + ESA, we observed a significant positive correlation between S100A9 and duration of LEN or LEN + ESA response (r=-0.418, p=0.03). In vitro, 1µM LEN significantly reduced suppression of Epo elaboration by S100A9 or TNFα in HepG2 cells. Magnitude of LEN's improvement in EPO elaboration was greater for S100A9 vs TNF (43% vs 17%, p=0.08).Moreover, in peripheral blood mononuclear cells (PBMC) from MDS patients, 1µM LEN significantly decreased steady state S100A9 generation, and LPS-induced TNFα elaboration.

Conclusion: These findings demonstrate that S100A9 and its NFκ-B transcriptional target, TNFα, directly suppress Epo elaboration and endocrine response to anemia in MDS. These inflammatory proteins may serve as rational biomarkers for response to lenalidomide and/or ESA treatment and merit prospective validation. Therapeutic strategies that either neutralize S100A9 or suppress elaboration of this alarmin may improve erythropoiesis in MDS by restoring Epo response and suppressing innate immune effectors.