Red Cells Distribution Width and Prognosis in Myelofibrosis Patients Treated with Ruxolitinib

Introduction

Myelofibrosis (MF) is a BCR::ABL1-negative myeloproliferative neoplasm characterized primarily by genetic abnormalities and an aberrant production of pro-inflammatory cytokines (including IL-2, IL-6, TNF-α, and TGF-β among others) released from pathological hematopoietic stem cells leading to the deposition of reactive fibrosis in the bone marrow (BM), at the same time the development of extramedullary hematopoiesis, mainly in the spleen and liver, and finally the production of red blood cells (RBC) of often bizarre shape, i.e. dacryocytes. Consequently, increased variability in RBC volume, so-called anisocytosis, is a hallmark of MF.

A small number of studies have already demonstrated that quantitative measurement of anisocytosis in terms of red cells distribution width (RDW) may have a negative prognostic role in newly diagnosed MF patients.

Aim

In this study we further evaluate the prognostic impact of RDW in a series of patients with primary (PMF) and secondary myelofibrosis (SMF) at diagnosis and during treatment with ruxolitinib (RUX).

Methods

Between October 2014 and May 2023, the RDW of 49 consecutive patients with PMF or SMF (32 males and 17 females; median age at RUX initiation, 70.5 years) were retrospectively assessed at MF diagnosis, at RUX start and after 3, 6, 12, 24 and 36 months from therapy initiation. All patients were treated with RUX according to current indications, requiring an IPSS risk of at least intermediate-1.

Red cells distribution width was routinely reported by automated cell counters as part of RBC count, with an upper limit of normal of 14.5% used as a cut-off in our laboratory.

At the time of RDW evaluation, complete blood cells count and spleen size [expressed as the distance from the left costal margin (BCM) in cm] were available for all patients.

Results

Among 49 patients, 24 had PMF (10 prefibrotic and 14 overt fibrotic) and 25 SMF [15 had post-polycythemia vera (PPV)- and 10 post-essential thrombocythemia (PET)-MF]. The JAK2V617F mutation was detected in 38 (77.6%) cases, CALR mutations in 8 (16.3%), and MPL in 1 (2%), while 2 patients (4.1%) were “triple-negative”.

Elevated RDW values were present in all but 3 patients at MF diagnosis with a median RDW of 18.7% (range, 13.3% - 25.7%). RDW was higher in patients with splenomegaly ( p=0.04) and with JAK2V617F mutation ( p=0.09). Furthermore, looking specifically at JAK2V617F-mutated cases, the greater the RDW values, the higher the JAK2V617F allele burden. Interestingly, a higher RDW value at diagnosis also correlated with a shorter time from MF diagnosis to RUX start (-3.5 months per one RDW unit, 95%CI: -7.5 to +0.4; p=0.08).

We observed a modest (<1%) increase in RDW during the first 6 months of RUX treatment. In a multivariable random-intercept model that considers all time points and contains the covariates time, RUX dose, and hemoglobin (Hb), we also observed a clear decrease in RDW with increasing Hb (slope: -0.35% per g/dL of Hb, 95%CI: -0.56 to -0.14; p=0.001).

Median overall survival (OS) from MF diagnosis was 6.5 years with a median duration of RUX treatment of 33.3 months. The median RDW at diagnosis of 18.7% was used as a cut-off to identify two subgroups of patients [Group 1 (N=22): RDW 18.7% to 25.7%; Group 2 (N=23): RDW 13.3% to 18.6%] showing a difference in OS [Group 1 vs. 2: Hazard Ratio (HR) 2.74, 95%CI: 1.13 to 6.65; p=0.03] ( Figure 1); the median OS was 8.5 (14 deaths) and 11.8 (9 deaths) years for the Group 1 and 2, respectively. Using continuous RDW at diagnosis, the crude HR was 1.21 per one RDW unit (95%CI: 1.06 to 1.37; p=0.004). In a Cox model adjusted for gender, age, and Hb, the HR was 1.13 (95%CI: 0.99 to 1.29; p=0.06). Notably, RDW at diagnosis was associated with a 35% increase in infectious disease mortality in a crude analysis (HR 1.35, 95%CI: 1.07 to 1.71, 8 deaths; p=0.01) and a 27% increase (HR 1.27, 95%CI: 0.99 to 1.63; p=0.07) in the adjusted analysis.

Conclusions

Red cells distribution width may represent a good integrative measure of several disease processes and also has prognostic significance at the time of MF diagnosis and during RUX treatment. Even though our preliminary observations are limited by the small number of patients and need to be further confirmed by larger studies, we can speculate that RDW may help in the rapid detection of MF patients with a poor prognosis, which can only partially be improved from currently available targeted therapies with JAK inhibitors.