Effects of Long Term Imatinib on Bone Mineral Density In Patients with Chronic Myelogenous Leukemia (CML) or Gastrointestinal Stromal Tumor (GIST)

Abstract 2276

Imatinib (IM) inhibits several disease-associated tyrosine kinases, including BCR-ABL in CML, C-KIT in GIST, and PDGFR α and β in patients (pts) with certain myeloproliferative disorders and dermatofibrosarcoma protuberans, respectively. We have previously shown that most pts taking IM have low levels of osteocalcin, a marker of bone formation, and serum N-telopeptide of collagen cross links (sNTX), a marker of bone resorption (Berman, 2006). We postulated that the decrease in bone turnover was, in part, related to the fact that IM inhibits PDGFR α and β which are important for osteoblast and osteoclast function, as well as M-CSF activation of C-FMS which is needed for osteoclast maturation from monocytic precursors (Dewar, 2005). In order to determine whether these findings had a clinical impact on pts taking IM for either CML or GIST, we undertook a prospective analysis of BMD in pts already on the drug. Patients with known metabolic bone disease or pts taking supplemental calcium or biphosphonates were excluded. At time of study entry, we examined fasting serum calcium, phosphate, parathyroid hormone (PTH), vitamin D (25 and diOH 1,25), osteocalcin, bone alkaline phosphatase (BAP), urinary phosphate excretion, and serum and urinary NTX. Tests were repeated every 3 months for 2 years. Bone densitometry was performed at time of study entry and then at 12 and 24 months using a GE Lunar densitometer. A significant change in BMD was defined as a decrease of > 3.6% at the lumbar spine, > 3.8% at the femoral neck, and > 2.4% at the hip. Nineteen adults (9 M/10F; 14 with CML)), median age 52 (range 24–73), completed the study. Nine of the 19 pts (47%) had a significant decrease in BMD in at least one site; this included 2 patients with a significant decrease at the femoral neck, 4 patients with a decrease at the hip, and 3 pts with a decrease at both femoral neck and hip. Four pts (21%) had a significant increase in BMD: 2 at the lumbar spine, and 2 at the femoral neck and hip. Six pts (31%) had no change in BMD. The median time on IM before study entry was 31 months (range 1–71) for all 19 pts. For the groups with a decrease, increase, or no change in BMD, the median times on treatment were 27, 3, and 54 months. Serum osteocalcin levels were consistently low in all 3 groups with a median of 2.7 ng/ml (range 2.7–3.4) after 24 months on study; 7 pts (37%) had no measurable amount of osteocalcin on at least one occasion. In all 3 groups, BAP was in the low- normal range, serum and urine NTX were normal, and PTH levels were consistent with vitamin D levels. Although numbers are small, bone loss occurred at sites consisting of cortical bone (hip and femoral neck), and not at the site of predominantly trabecular bone (lumbar spine). In summary, BMD may decrease in ∼50 % of patients taking IM. Osteocalcin levels remained low in nearly all pts while serum NTX levels were normal, suggesting that IM inhibits osteoblast function to a greater degree than osteoclast function. It remains to be determined which pts need serial BMD testing, and whether bisphosphonates or other agents might improve BMD in this setting.