BACKGROUND: Bone marrow fibrosis in myelofibrosis (MF) is the result of a complex and poorly understood interaction between megakaryocytes, fibroblasts, endothelial cells, cytokines and marrow stroma. Preclinical studies support a pathobiological role of TGF-β. It is overexpressed by megakaryocytes in MF and the TGF-β signature is upregulated and has been targeted in MF animal models. However TGF-β is a pleiotropic cytokine implicated in many cellular processes. CCN proteins are a group of 6 matricellular proteins important in fibrotic diseases and injury repair. CCN1 (CYR61) and CCN2 (CTGF) are transcriptionally activated by mitogenic growth factors such as PDGF, FGF2 and TGF-β. Recombinant CCN2 induces differentiation of human bone marrow mesenchymal stem cells into fibroblasts (Lee et al. JCI 2010). Levels of CCN2 in biologic fluids correlate with severity of fibrosis in scleroderma, liver cirrhosis and nephropathy. Given the proven role of CCN2 as a measurable serum biomarker in pro-fibrotic diseases and a downstream effector of TGF-β, in this retrospective study we examined CCN2 expression in myelofibrosis. We studied its correlation with clinical response following allogeneic stem cell transplant (ASCT), the only therapeutic modality to date which can consistently reverse fibrosis.

METHODS: Patients diagnosed with MF at our institution from 1998 to 2015 were identified by diagnostic code (IRB#2013-0896). Bone marrow (BM) specimens at diagnosis and 1 year following ASCT were retrieved. Staging bone marrows from lymphoma patients, read as normal, were used as negative controls. CCN2 localization and expression was assessed by IHC (Abcam 5097). The slides were scored from 0-100% for megakaryocyte cytoplasmic staining by a blinded hematopathologist. RNA was extracted and reverse transcribed from frozen fixed paraffin embedded (FFPE) BM samples using the Qiagen RNeasy FFPE RNA Purification Kit. Assays for CCN1, CCN2 and CCN3 (NOV) and endogenous controls RPLPO and GAPDH were performed using Taqman quantitative PCR assays. mRNA data were analyzed using the software package DataAssist (v3.01; Life Technologies).

RESULTS: CCN2 expression is upregulated in myelofibrosis compared to healthy bone marrow. mRNA analysis showed a 27 fold increase (p=0.19) in CCN2 mRNA expression in patients with MF when compared to healthy controls. IHC data also showed increased expression in megakaryocytes of MF patients compared to controls (63% vs 40%, p=0.28). Exploratory analysis showed a 36 fold increase in CCN2 mRNA expression in JAK2 negative (n=3) MF patients compared to JAK2 positive (n=7) patients (p=0.06), suggesting expression may depend on the molecular profile. CCN2 expression in the bone marrow of myelofibrosis patients is significantly downregulated at 1 year after allogeneic stem cell transplant. 13 pre-treatment cases of MF were compared with 6 post-transplant cases. The mean percentage of megakaryocyte with cytoplasmic expression of CCN2 by IHC in MF at diagnosis was 63%, compared to 22% in post-transplant specimens, (p= 0.01). mRNA extracted from the post-transplant marrows showed 0.01 fold expression of CCN2 compared to pre-transplant (p= 0.18). This decrease in CCN2 correlated with clinical and pathologic resolution of disease. The average DIPSS score pre-transplant was 2 (range 0-5) which improved to 0 for all patients post-transplant. The mean BM cellularity pre-transplant was 82% (range 50%-95%) and post-transplant was 48% (range 35%-65%). The average decrease in cellularity after transplant was 29%. Reticulin fibrosis ranged from grade 2-3 in the pre-transplant MF bone marrow samples which improved to grade 0-1 post-transplant.

CONCLUSIONS: CCN2 is a downstream effector of TGF-β and a measurable biomarker of fibrosis in fibroproliferative diseases. We show that CCN2 expression in myelofibrosis decreases significantly after ASCT, suggesting its role as a biomarker of fibrosis in this disease. This paralleled clinical and pathologic resolution of disease. This will need to be validated in a larger number of paired patient samples where in addition to bone marrow expression, serum CCN2 levels will be measured by ELISA.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.