Proton Pump Inhibitors Increase Local Reparative Bone Mass around Calcium Phosphatecement Implants in Aged Osteopenic Rats

INTRODUCTION: Successful integration of osseous implants with host bone tissue can be adversely affected by pathological bone conditions such as osteoporosis, which results in osteopenia due to a perturbation of the balance between the bone resorption and bone formation processes. . Bone resorption is accomplished by osteoclasts which create an acidic environment by secreting protons through vacuolar proton pumps (VATPases). Because the osteoclast proton pump is unique, its function can be affected by pharmacological agents, i.e., proton pump inhibitors (PPIs), which intervene with the osteoclast resorptive activity and therefore result in reduced bone resorption. Although extensive data describing the systemic effects of osteoclast proton pump inhibitors on bone resorption are available, no one has proposed using these compounds as a means of stimulating local increase in new reparative bone formation around resorbable implants such as calcium phosphate (CAP) cements. We have incorporated a specific osteoclast proton pump inhibitor, Bafilomycin A1 (BAF), into a resorbable CAP cement to determine whether we can stimulate an increase in reparative bone mass around the material and decrease resorption of the newly formed bone at the defect site in a pathological bone tissue environment. The animals we chose for this study were aged ovariectomized (ovx) rats, which have become osteopenic as a result of both old age and ovariectomy. Because the aged ovx rat has been validated by the FDA as a clinically relevant model of human postmenopausal bone loss, the results from this study can provide important information about the feasibility of using PPI-modified resorbable implants for stimulating regenerative bone growth in postmenopausal patients. METHODS: Dicalcium phosphate anhydrous (DCPA), dimethyl sulfoxide (DMSO), and Bafilomycin A1 (BAF) were purchased from SigmaAldrich Inc. (Oakville, ON, Canada). Tetracalcium phosphate (TTCP) was purchased from Clarkson Chromatography Products Inc. (South Williamsport, PA, USA). Disodium hydrogen phosphate (Na2HPO4) and sodium dihydrogen phosphate (NaH2PO4) were purchased from BDH Inc. (Toronto, ON, Canada) and used to prepare a phosphate solution (pH 7.4). BAF was dissolved in DMSO and incorporated into the phosphate solution. CAP precursor powder consisted of an equimolar mixture of DCPA and TTCP. Control and BAF-containing cement pastes were prepared by mixing the powder and the liquid at a ratio of 2.0 wt/wt. Cement rods (1.9 mm diam. x 2.3 mm height) were shaped by allowing aliquots of packed CAP pastes to set in custom-made Teflon formers at 37°C and 100% humidity. 10-month old ovx virgin Brown Norway rats were purchased from the National Institute on Aging (Bethesda, MD, U.S.), housed individually in lightand temperature-controlled rooms, and fed a restricted diet (6 pellets/day). These animals were kept for additional 4 months following the ovariectomy. Non-ovx virgin female Brown Norways of the same age at the implantation time (14 months old) as the experimental animals and young virgin females (3 months old) were used as controls and comparison groups for the aged ovx rats for evaluating the implant integration and reparative bone formation/resorption. Six animals from each group were used for each time period. Bone defects (2.3 mm diam.) were drilled in the distal femurs of the rats and the pre-set rods were placed into the defects. Every rat received a control rod (CAP) in one femur and a rod containing BAF (CAP-BAF) in the other femur. After desired time in vivo, the femurs were removed, fixed, decalcified, and embedded in low-melting-point paraffin. Qualitative evaluation of the long-term (7 days up to 4 months in vivo) local effect of BAF on new reparative bone formation and resorption around the implants was done using histological techniques. Serial sections perpendicular to the long axis of the implant were obtained using Spencer 820 microtome. The sections were stained with Masson’s Trichrome, and tested for the presence of tartrate resistant acid phosphatase (TRAP) enzyme. RESULTS SECTION: In control specimens, “cutting cones” formed by multinucleated cells were resorbing the rods as early as at 10 days. These multinucleated cells showed positive staining for TRAP that putatively identified them as osteoclasts. The cements containing BAF did not show this resorption pattern in the early implantation stages, where TRAP-positive osteoclasts were present at the material surface, but did not resorb the material until after 1 month. Thus, the appearance of the control cements was progressively distorted with time (Figs. 1A & 2A), while CAP-BAF cements resisted osteoclastic resorption (i.e., retained their cylindrical shape) for a longer period of time (Figs. 1B & 2B). In addition, all cements containing BAF (Figs. 1B & 2B) were consistently surrounded by more reparative bone than the control cements (Figs. 1A & 2A) throughout the implantation periods studied. New bone that formed around BAF-modified cements appeared denser, as judged by histology, which could be attributed to the decreased resorptive activity of the osteoclasts due to the presence of the osteoclast proton pump inhibitor in the implant. Indeed, consistent differences in the amount and density of reparative bone surrounding the control and BAF-modified cements remained evident even at 2 months in vivo (compare Figs. 2A & 2B), at which time the control cement was associated with only a few disconnected trabeculae whereas the cement with BAF was surrounded by a solid ring of bone. DISCUSSION: The high degree of osteoclastic activity and the resorption rate of the control cements were decreased in the case of the CAP-BAF cements due to the inhibition of proton pumps by the BAF. This also resulted in the maintenance of the original shape and size of the CAP-BAF cement rods. We believe that BAF slowed down local resorption of the new bone by leaching from the cement matrix into the surrounding periimplant environment. In conclusion, osteoclast proton pump inhibitors, when incorporated into resorbable implants, can be used to stimulate local reparative bone mass and slow down the resorption of newly formed bone in osteopenic rats. Furthermore, introduction of the PPI into the cement matrix also allows control over implant resorption. REFERENCES: 1. Chatterjee D et al 1992, J Exp Biol 172:193-204. 2. Farina C, Gagliardi S 1999, Drug Discovery Today 4(4):163-172. 3. Marks SC Jr., Sundquist KT 1995, Eur J Oral Sci 103(4):231-5. 4. Herslof M et al 1994, U.S. Patent 5354773. Figure 1. Representative transverse section of an ovx rat femur containing pre-set calcium phosphate cement rod at 10 days in vivo: (A) Control CAP, (B) CAP-BAF. Masson’s Trichrome stain. FW= 7.5 mm. N=6. Figure 2. Representative transverse section of a rat femur containing pre-set calcium phosphate cement rod at 60 days in vivo: (A) Control CAP, (B) CAP-BAF. Masson’s Trichrome stain. FW= 7.5 mm. N=6. A