Changes in Activated Bone Marrow Macrophages and Mast Cells in Patients with Myelofibrosis Following Ruxolitinib Therapy

Background: Macrophages (MAC) are key regulators of malignant progression in solid tumors by promoting tumor cell invasion, migration and angiogenesis. Only few studies have investigated the role of activated MAC phenotypes, such as M1 and M2, in the bone marrow (BM) microenvironment of myelofibrosis (MF). MAC are generally increased in the BM of patients (pts) with MF and recent studies have also identified mast cells (MC) as playing an important role in regulating the underlying inflammatory process in MF. With regard to the substantial improvement in the mostly cytokine-driven constitutional symptoms following JAK inhibition therapy, we have investigated the effect of ruxolitinib (RUX) on BM MAC subtypes and MC and their association with BM fibrosis.

Methods: A total of 63 pts with high-risk MF were included in this analysis. All pts had both baseline (BL) and sequential BM trephine biopsy at 24 months (mo) following RUX therapy. Using specific immunohistochemical stains, we analyzed assorted MAC markers, including CD68 and CD163. CD68 is a pan-MAC marker that recognizes both M1 and anti-inflammatory M2 subtypes. CD163 is a scavenger receptor upregulated by MAC within an anti-inflammatory environment and regarded as a highly specific monocyte/macrophage marker for the M2 subtype. Anti-Mast Cell Tryptase was applied as specific antibody to identify BM MC. All sections were stained following routine procedures and quantification of positivity was performed for each marker by consensus after independent review by 3 pathologists. Cytokine expression levels (TNF-alpha, MIP-1-alpha, IgE) at week 4 and 24 mo following RUX therapy were available for a subset of 23 and 16 pts. Individual changes were categorized as increase, stable, or decrease and correlated to the degree of WHO BM fibrosis grade and hematological features.

Results: At BL, 81.0% of MF pts presented with an advanced fibrotic stage of disease (WHO grade 2 or 3). Grade of BM fibrosis significantly correlated with the frequency of CD68+ and CD168+ MAC. Furthermore, advanced fibrosis was associated with a higher frequency of BM MC. Following RUX treatment, 14.3% of cases showed an improvement in BM fibrosis, while 58.7% showed stabilization. RUX induced in 48.3% of cases a significant decrease in the overall amount of BM CD68+ MAC, whereas a further increase was observed in only 6.9%. Similar results were obtained for the specific CD163+ anti-inflammatory M2 subtype. Post-RUX a significant reduction of this cell lineage was seen in 47.6%, while 12.7% of cases revealed a further increase. Improvement in BM fibrosis was highly correlated with an overall reduction of CD68+ MAC, and in particular with modulation of the CD163+ M2 subtype. RUX therapy induced a profound reduction in the expression of associated cytokines such as TNF-alpha and MIP-1-alpha, both at 4 weeks and 24 mo. Frequency of BM MC was reduced in 26.3% of pts during therapy, however, in 49.1% therapy showed no change. Pts with improvement in BM fibrosis revealed in most cases lower frequencies of MC, but this association did not reach statistical significance. In contrast, expression levels of IgE were strongly reduced in almost all pts at week 4 and 24 mo. The amount of MAC or MC did not correlate with BL hematological or clinical parameters such as BL spleen size, anemia or platelet counts. Following RUX, decreases in M1 and M2 MAC were associated with changes in hemoglobin levels and spleen size reduction.

Conclusions: Our results significantly extend previous observations on the role of the BM microenvironment in MF. RUX treatment meaningfully and directionally impacts the amount of activated anti-inflammatory MAC in BM of MF patients. Overall, our data advocate the strong disease modulation capacity of anti-JAK therapy.