Macrophage frequency in the bone marrow correlates with morphologic subtype of myeloproliferative neoplasm

Abstract

Bone marrow (BM) fibrosis in myeloproliferative neoplasms (MPNs) is associated with a poor prognosis. The development of myelofibrosis and differentiation of mesenchymal stromal cells to profibrotic myofibroblasts depends on macrophages. Here, we compared macrophage frequencies in BM biopsies of MPN patients and controls (patients with non-neoplastic processes), including primary myelofibrosis (PMF, n = 18), essential thrombocythemia (ET, n = 14), polycythemia vera (PV, n = 12), and Philadelphia chromosome-positive chronic myeloid leukemia (CML, n = 9). In PMF, CD68-positive macrophages were greatly increased compared to CML (p = 0.017) and control BM (p < 0.001). Similar findings were observed by CD163 staining (PMF vs. CML: p = 0.017; PMF vs. control: p < 0.001). Moreover, CD68-positive macrophages were increased in PV compared with ET (p = 0.009) and reactive cases (p < 0.001). PMF had higher frequencies of macrophages than PV (CD68: p < 0.001; CD163: p < 0.001) and ET (CD68: p < 0.001; CD163: p < 0.001). CD163 and CD68 were often co-expressed in macrophages with stellate morphology in Philadelphia chromosome-negative MPN, resulting in a sponge-like reticular network that may be a key regulator of unbalanced hematopoiesis in the BM space and may explain differences in cellularity and clinical course.

Keywords: Bone marrow fibrosis; Immunohistochemistry; Macrophages; Myeloproliferative neoplasm.

Fig. 1

Macrophage frequency and distribution in MPN and reactive bone marrow biopsies. Representative photomicrographs showing CD68-expressing macrophages by IHC (upper) and CD163 staining (lower). Percentages of CD68-positive cells determined by automatic cell counting in the upper row cases from left to right: PMF (30%), ET (2%), PV (18%), CML (2%), and control (2%). Percentages of CD163-positive cells in the lower row paired samples from left to right: PMF (38%), ET (0%), PV (23%), CML (2%), and control (2%). Evenly dispersed CD68- and CD163-expressing macrophages, with stellate morphology, are increased in Ph− MPN (especially in PMF and PV) and show a more ovaloid morphology in CML (× 400 magnification)

Fig. 2

Correlation of macrophage frequency with type of MPN. Scatter plot, CD68 (A) and CD163 (B) frequencies in biopsies of MPN patients (PMF, PV, ET, CML) and reactive BM controls. CML and reactive BM contain few macrophages, while PMF biopsies contain higher frequencies of CD68- and CD163-positive macrophages (% per nucleated cells). PV was associated with a significant higher macrophage frequency than CML or ET. Brackets indicate significant differences between diagnostic subtypes (p < 0.05)

Fig. 3

Correlation of CD68-positive and CD163-expressing macrophage frequency. Macrophage frequencies determined by CD68 and CD163 IHC staining (immunopositive cells per all nucleated cells, QuPath analysis)

Fig. 4

Immunofluorescence co-staining of CD68 and CD163 in MPN biopsies. Multispectral microscopy reveals frequent co-localization of CD68- (green) and CD163 (red)-positive macrophages, nuclei highlighted in DAPI (blue), representative images (× 200) (white arrow: CD68/CD163 double-positive macrophages)

Fig. 5

Myelofibrosis (MF) grade and macrophage frequency. Scatter plot comparing MF grade and macrophage frequency using CD68 (A) and CD163 (B) automatic cell counts (% positive cells by IHC per nucleated cells, QuPath analysis) in BM trephines. MF grade was determined using a four-tiered scale [16]: MF grade with no increase in reticulin fibers (grade 0), loose increase of reticulin (1), more diffuse and dense increase in reticulin and beginning collagen fibers (2), and dense reticulin and collagen fibrosis (3)