Reconstructing Skeletal Homeostasis through Allogeneic Hematopoietic Stem Cell Transplantation in Myelofibrosis

Myeloproliferative neoplasm-associated myelofibrosis (MF) is a clonal stem cell disease characterized by pronounced bone marrow fibrosis associated with extramedullary hematopoiesis and splenomegaly. Allogeneic hematopoietic stem cell transplantation (allo-HSCT) represents the only curative treatment leading to MF regression. The involvement of skeletal homeostasis in the process of MF regression after allo-HSCT has remained unclear.

To delineate the dynamics leading to the resolution of MF after allo-HSCT we included 66 patients in this prospective cohort study of whom 40 (61%), 16 (24%), and 10 (15%) had primary myelofibrosis, essential thrombocythemia, and polycythemia vera, respectively. Most patients showed advanced MF grade with 73% of patients having MF-3 at time of allo-HSCT. To understand the changes to the bone and the bone marrow we collected bone marrow trephine biopsies before and after transplantation; and blood and urine samples before and on day 30, 100 and 365 after transplantation. In addition, we assessed the bone status by dual-energy X-ray absorptiometry (DXA) and high-resolution peripheral quantitative computed tomography (HR-pQCT). Measurements were compared to a bone-healthy cohort by propensity score matching.

DXA measurements showed highly increased bone mineral density (BMD) at the lumbar spine compared to after allo-HSCT (T-score 1.4 ± 1.8 vs. 0.0 ± 1.2, P < 0.0001), often in the range of osteosclerosis. Similar results with even larger differences were observed at the hip (T-score, 1.4 ± 1.9 vs. -0.6 ± 0.8, P < 0.0001). After allo-HSCT patients showed no significant differences compared with the control cohort.

Interestingly, there was no difference in bone microarchitecture and BMD visible by HR-PQCT neither in radial nor tibial measurement sites. The three-dimensional structural parameters total volumetric BMD, bone volume to tissue volume and cortical thickness showed no decrease after transplantation and were comparable to the controls.

Biopsies obtained before allo-HSCT showed markedly increased extracellular matrix (ECM) deposition and decreased adipocytes. There was an elevation in trabecular number (Tb.N), a trend toward a higher ratio of osteoid volume to bone volume, while the trabecular thickness (Tb.Th) showed normal values. After allo-HSCT we observed normalization of ECM and a restoration of bone marrow adipose tissue. Tb.N showed a decrease while Tb.Th remained unchanged, both now comparable to the control cohort.

Quantitative backscattered electron imaging (qBEI) of the embedded trephine biopsies was performed to assess calcium content. Using qBEI, we were able to demonstrate a disturbance in bone mineralization in MF patients, as indicated by a decreased mean calcium content compared to controls. These changes, while alleviated, remained visible after allo-HSCT.

Blood and urine samples were analyzed to estimate bone and mineral metabolism. Patients before transplantation showed highly elevated levels of parathyroid hormone with a steady decrease following transplantation. Interestingly, DPD crosslinks in urine showed a spike in the first year after transplantation indicating bone resorption.

In conclusion, we outlined the landscape of changes to the bone of patients with MF undergoing allo-HSCT. There is a clear indication of high bone turnover in these patients suffering from a clonal hematopoietic disease, which can be partially rescued by allo-HSCT. The molecular and cellular dynamics of these processes remain to be understood.