Clinical and Genetic Factors Underlying Acute Bone Mineral Density Loss by 100 Days after Blood and Marrow Transplantation: A Potential Early Regimen-Related Complication

Abstract

Accelerated bone mineral density (BMD) loss after blood and marrow transplantation (BMT) has been described, with corticosteroid dose as the most important risk factor. Loss of BMD exponentially increases fracture risk, where a 10–15% decrease in BMD approximately doubles the risk. Our preliminary study [BMT 41:393–398, 2008] demonstrated that BMD loss frequently occurs even among patients with little to no steroid exposure. In the general population, common genetic variations are estimated to contribute as much as 50–80% of the phenotypic variation in bone-related traits, however genetic factors have not been examined in relation to BMD loss post-BMT. In the current study, we prospectively measured BMD by dual energy X-ray absorptiometry (DXA) scans at the lumbar spine and dual femur on all adult (≥18 yrs) first autologous or allogeneic BMT patients at our center. Pre- and post-BMT serum levels of 25(OH)D, parathyroid hormone (PTH) and Ca++ were collected within 2 weeks of each DXA scan. From 1/2006 and 4/2008, 151/157 (96%) BMT patients had a DXA scan a median 22 days pre-BMT, and 106/151 had a second DXA scan a median 100 days post-BMT. Reasons for not having a second DXA were early mortality (n=32) or medically unfit (n=13). Median age was 51 yrs, 61% were male, and 92% were White. The diseases included ALL (5%), AML (23%), lymphoma (39%), MDS/aplastic anemia (9%), and myeloma/amyloidosis (25%). BMT type was 57% autologous and 43% allogeneic, with 86% receiving blood stem cells. Conditioning and treatment regimens were myeloablative (74%) or reduced intensity (26%). While BMD for the majority remained in the normal range, there was a significantly accelerated BMD loss in most patients. After standardizing the time between pre- and post-BMT DXAs to 100 days, the median BMD change was −0.031 g/cm2 at the spine and −0.043 g/cm2 at the femur, which equates to a median annual loss rate of 9.2% and 14.4%, respectively. In men and premenopausal women, the average annual BMD loss rate is 0.2–0.4% per year, and increases after menopause to 1–2% per year. When compared to the expected BMD loss rates among healthy adults and adjusted for gender and age, the loss seen in BMT patients measured in a 4 month period was equivalent to aging their bones by approximately 12 yrs at the spine and 9 yrs at the femur. Serum levels of 25(OH) D, PTH, or Ca++ were not associated with BMD measured pre- or post-BMT nor with BMD loss. Multivariate analysis demonstrated the following significant risk factors for BMD loss at the spine: steroid dose, baseline BMD, and conditioning regimen; and at the femur: steroid dose, baseline BMD, age, gender, BMI and stem cell type. The multivariate models did not demonstrate a difference between autologous and allogeneic BMT. To estimate the performance of these clinical factors in explaining the variance in BMD loss, we calculated R2 of the best multivariate models. Although statistically significant (P<0.0001), our best model explained only 17% of the variance in BMD loss at the spine and 44% at the femur. We are currently analyzing 170 genetic polymorphisms selected from 46 candidate genes from 6 pathways: vitamin D-related genes, cytokines and their receptors, bone matrix genes, estrogen-related genes, PTH and steroid-related genes, and other genes. Gene-based and pathway-based approaches will be used to identify genetic factors that play an important role in BMD loss related to BMT. Finally, a complete model integrating clinical and genetic factors will be built to predict risk of bone loss, and early interventions can therefore be tailored to patients at high risk. It is clear that accelerated BMD loss occurs early after BMT which increases the risk of osteopenia, osteoporosis and fractures at an earlier age in the BMT population than in the general population. Thus, BMD loss could be considered a regimen-related toxicity due to its close proximity to and association with BMT.