Abstract
Introduction:Chronic myelomonocytic leukemia (CMML) is a genetically heterogeneous myeloid neoplasm characterized by the presence of both dysplastic and proliferative features and highly variable clinical outcome. A CMML specific prognostic system (CPSS) has been developed that stratifies patients in to low, intermediate and high risk groups based on WHO subtype, FAB subtype, transfusion dependent anemia, and karyotype. Somatic mutations and DNA methylation patterns can increase prognostic precision, but fail to explain a large part of the clinical variation, suggesting that additional variables, including comorbidities, may be major determinants of overall survival (OS) in CMML.
Methods: We retrospectively identified CMML patients diagnosed between 1996 and 2017 at the Huntsman Cancer Hospital, University of Utah, using ICD codes, tumor registry data and electronic medical records. For all patients a diagnosis of CMML was confirmed based on 2008 WHO diagnostic criteria. Data on comorbidities at the time of diagnosis were obtained by search of electronic medical records using a customized rule based algorithm utilizing linguimatics text mining software (Natural language processing). The comorbidities were scored and categorized as per previously published reports: low, intermediate and high risk groups for MDS Comorbidity Index (MDS-CI) and low, mild, moderate/high (moderate and high included in the same group due to small number of patients) for the Charlson Comorbidity Index (CCI). Continuous variables were transformed into categorical variables, based on cutoffs used in previously published studies.
Univariate analysis was performed using the Cox proportional hazards model for categories: MDS-CI (low, intermediate and high) and CCI (low, mild, moderate/high). Other variables analyzed included age (<70 or >70 years), sex (male or female), hemoglobin (<10 gm/dL or >10 gm/dL), platelet count (<100k/uL or >100k/uL), WHO subtype (CMML-0, CMML-1 and CMML-2), FAB subtype (CMML-MD or CMML-MP), karyotype (low, intermediate and high risk) and treatment with hypomethylating agents (yes or no). Kaplan-Meier methods were used for plotting OS. All analysis was performed using R statistical programming software version 3.2.1 (The R Foundation for Statistical Computing, Vienna, Austria). Results shown are censored at the time of allogeneic stem cell transplant. For OS the "Low" category is reference and the p-values are for comparison to this category using the Cox model.
Results: We identified 94 patients with confirmed diagnosis of CMML. The median age was 76 (range 33-91 years) and 61 patients were men (65%). Fifty-five (58.5%), 34 (36.2%) and 5 (5.3%) patients were categorized as MDS-CI low, intermediate and high risk respectively. Sixty-two (66%), 26 (27.6%) and 6 (6.4%) were categorized as low, mild and moderate/high CCI risk. Hazard ratios (HR) for MDS-CI risk categories were: intermediate=1.26 (95% CI 0.71 to 2.23; p=0.425) and high risk=2.22 (95% CI 0.86-5.75); p=0.101). HR for CCI risk categories were: mild=1.01 (95% CI 0.56-1.82; p=0.964) moderate/high=4.18 (95% CI 1.57 to 11.10; p=0.004). HR for other variables are shown in Table 1. Kaplan-Meier curve representing the OS of the entire cohort categorized according to CCI and MDS-CI risk categoriesis shown in Figure 1. Estimated median survival for MDS-CI low, intermediate and high is 36, 36, and 23 months respectively. Median survival for CCI-CI low, mild, moderate/high risk categories was 36, 33, and 10 months respectively (Figure 1).
Conclusions: High risk CCI and MDS-CI category patients are at markedly higher risk of death, suggesting that co-morbidities are major host-related determinant of OS in CMML. Given the association of clonal hematopoiesis of indeterminate potential (CHIP) with coronary heart disease (Jaiswal et al. N Engl J Med 2017; 377:111-121) and the fact that CHIP genes such as TET2are frequently mutated in CMML, it is conceivable that CMML causally contributes to comorbidities. Somatic mutation data are being collected for inclusion in a multivariate model that will be presented at the conference.
Shami:JSK Therapeutics: Employment, Equity Ownership, Membership on an entity's Board of Directors or advisory committees; Lone Star Biotherapies: Equity Ownership; Pfizer: Consultancy; Baston Biologics Company: Membership on an entity's Board of Directors or advisory committees. Kovacsovics:Abbvie: Research Funding; Amgen: Honoraria, Research Funding. Deininger:Pfizer: Consultancy, Membership on an entity's Board of Directors or advisory committees; Blueprint: Consultancy.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal