P755: HYPERFERRITINEMIA IS A PREDICTIVE BIOMARKER OF POOR CLINICAL OUTCOMES IN CMML

Abstract

Background: CMML is a heterogenous disease exhibiting features innate to MPN and MDS. Increasing evidence supports a close interplay between systemic inflammation and risk of myeloid malignancies, notably for those with history of infection or autoimmune disease. CMML has been associated with inflammation and end-organ damage related to CKD and CVD. Analysis of gene signatures from CMML-derived monocytes has shown them to be highly proinflammatory. High ferritin may serve as a practical biomarker of disease activity to help identify pts at higher risk of poor outcomes. Aims: We aimed to identify whether hyperferritinemia correlated with worse outcomes in CMML and to characterize relevant baseline clinical and molecular features associated with this clinical phenotype. Methods: Retrospective data was collected from a database of pts with CMML treated at Moffitt Cancer Center. Pts were stratified in 2 cohorts based on ferritin levels (<1000 or ≥1000 ng/mL). Hyperferritinemia was defined as ferritin >1000 as seen at diagnosis or during follow-up. Kaplan–Meier was used to estimate OS. Cox regression was used for multivariate analysis. Results: Between August 1995 and October 2020 729 pts with CMML were identified. Median age at diagnosis was 71 (17-95). Out of 571 pts with available ferritin levels 29% (n=168) developed hyperferritinemia vs 71% (n=403) who did not. mOS was 32.4 mos (95%CI 30-35 mos). Pts with higher ferritin tended to present with CMML-2 (p=0.001) and harbor a proliferative phenotype (p=0.01). They presented with higher marrow cellularity (mean 83%, p =0.08), PLT (mean 177k, p= 0.038), and lower Hb (mean 9.5, p<0.05). There was no association with % circulating IMC, monocytes, WBC or ANC at baseline. Hyperferritinemia was associated with more profound fibrosis (p=0.007), cytopenias (p<0.05), % peripheral blasts (p<0.05), RBC and PLT transfusion dependence (p<0.05). Pts with hyperferritinemia had higher risk disease per IPSS-R, CPSS and all CMML models (p<0.05); and had higher rates of AML transformation (p<0.05). Pts were also more likely to require treatment earlier (within 3 yrs of diagnosis) (p<0.05). ASXL1 (p=0.002), EZH2 (p=0.003), and SETBP1 (p=0.019) mutations were more common among pts who developed hyperferritinemia. Conversely, TET2 (p=0.001), CBL (p=0.028) and SRSF2 (p=0.003) mutations were less common. mOS for pts with hyperferritinemia was 23.9 mos (95%CI 19.9-27.9 mos), much lower than for those with ferritin <1000 (mOS 40.5 mos, 95%CI 35.4-45.5 mos) (p<0.05). In multivariate analysis, hyperferritinemia was a significant independent covariate for OS after adjusting for CPSS, transfusion dependence and disease phenotype (dysplastic vs proliferative) (HR= 0.69; 95%CI 0.53-0.89; p=0.005). Summary/Conclusion: Almost 1/3 of pts with CMML will develop hyperferritinemia. This is associated with more aggressive disease and rates of AML transformation leading to dismal outcomes. ASXL1, EZH2, and SETBP1 MTs confer a higher risk of hyperferritinemia. Our findings indicate that hyperferritinemia is an independent prognostic biomarker that may serve as a surrogate representative of disease biology and comorbidities in CMML.