Primary myelofibrosis (PMF) is characterized by ineffective erythropoiesis, an underlying inflammatory state and potentially, iron overload in the setting of red cell transfusions. The potential interaction between these clinical variables and assessment of their independent prognostic contribution, if any, is of interest within the context of well established prognostic parameters including the DIPSS-plus model and increased circulating levels of inflammatory cytokines (J Clin Oncol. 2011 Apr 1;29(10):1356–63). Consequently, we studied the relationship between circulating levels of hepcidin (a surrogate for inflammation, iron stores and ineffective erythropoiesis), ferritin (iron stores and inflammation), red cell transfusion need (ineffective erythropoiesis) and soluble inflammatory cytokine levels in PMF patients.
The diagnosis of PMF was according to WHO criteria. Inclusion in this study required availability of archived plasma, bone marrow biopsy, and cytogenetic information at the time of first referral. Plasma total hepcidin levels were measured by ELISA (USCN Life Sciences, Wuhan, China). Serum ferritin was measured as part of routine clinical assessment (normal range: males: 24–336 mcg/L; females: 11–307 mcg/L). Concentrations of plasma cytokines were analyzed in duplicate by using Multiplex Bead-Based Luminex technology (Invitrogen, Carlsbad, CA).
Two hundred and three PMF patients met the above stipulated criteria; 136 (67%) were males and the median age at referral was 63 years (range 17–83). Their dynamic international prognostic scoring system (DIPSS)-plus risk categorization was: low-risk 22 (11%), intermediate-1 26 (13%), intermediate-2 85 (42%) and high-risk 70 (34%).
Hepcidin levels (median 156279 pg/mL; range 8082–2088002) were strongly correlated with serum ferritin (n=146; median 207 mcg/L; range 14–5367) (r2=0.43, p <0.0001). Both were directly correlated with red cell transfusion need and age, and inversely with hemoglobin level (all p <0.05); however, there was no correlation with increased levels of inflammatory cytokines (>3 standard deviation above mean of normal controls for IL-2 and/ or IL-8) (p>0.05).
At a median follow up of 35 months, 104 (51%) deaths and 18 (9%) leukemic transformations were recorded. In univariate analysis, increased levels of hepcidin (HR=2.0; p=0.0006) and ferritin (HR=2.7; p<0.0001) were predictive of inferior overall survival and this significance was sustained during multivariable analysis that included either increased inflammatory cytokine levels (HR=3.2 and 2.7, respectively) or transfusion need (HR=1.7 and 2.2, respectively) as covariates. When all 4 variables (i.e. increased levels of hepcidin, ferritin, inflammatory cytokines and transfusion need) were considered together in the Cox model, only increased inflammatory cytokine levels retained their prognostic significance (HR=2.7; p=0.0046). Furthermore, neither increased ferritin nor increased hepcidin levels were prognostically significant for overall survival when tested either individually or together against the DIPSS-plus model (p>0.05).
In univariate analysis, increased hepcidin level, but not increased ferritin or transfusion need, was predictive for inferior leukemia-free survival (HR=2.8; p=0.04) and this significance was sustained in multivariable analysis that included either cytogenetic risk categories (very-high risk versus high-risk versus other) (HR=2.7; p=0.046) or thrombocytopenia (platelet count <100 × 109/L) (HR=2.7; p=0.047) as covariates.
Circulating levels of hepcidin and ferritin are highly correlated variables in PMF and appear to provide redundant prognostic information for overall survival in this setting. The lack of correlation with inflammatory cytokines suggests that ineffective erythropoiesis may be the common link for their apparent prognostic value noted during univariate analysis. In contrast to inflammatory cytokines, neither hepcidin or ferritin levels were useful as a complement to the DIPSS-plus prognostic model in the current analysis. The relationship between hepcidin and leukemia-free survival is intriguing however the pathogenetic mechanism remains obscure and needs confirmation in additional patients.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.