lable at ScienceDirect

Biomaterials 32 (2011) 5411e5416

Contents lists avai

Biomaterials

journal homepage: www.elsevier .com/locate/biomater ia ls

An injectable calcium phosphate cement for the local deliveryof paclitaxel to bone

Marco A. Lopez-Heredia a, G.J. Bernard Kamphuis a, Peter C. Thüne b, F. Cumhur Öner c,John A. Jansen a,*, X. Frank Walboomers a

aDepartment of Biomaterials, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlandsb Schuit Institute of Catalysis, Eindhoven University of Technology, Eindhoven, The NetherlandscDepartment of Orthopaedics, University Medical Center Utrecht, Utrecht, The Netherlands

a r t i c l e i n f o

Article history:Received 24 February 2011Accepted 5 April 2011Available online 6 May 2011

Keywords:Calcium phosphate cementBone repairChemotherapyDrug delivery

* Corresponding author. Tel.: þ31 (0) 243614920; fE-mail address: j.jansen@dent.umcn.nl (J.A. Jansen

0142-9612/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.biomaterials.2011.04.010

a b s t r a c t

Bone metastases are usually treated by surgical removal, fixation and chemotherapeutic treatment. Bonecement is used to fill the resection voids. The aim of this study was to develop a local drug deliverysystem using a calcium phosphate cement (CPC) as carrier for chemotherapeutic agents. CPC consisted ofalpha-tricalcium phosphate, calcium phosphate dibasic and precipitated hydroxyapatite powders anda 2% Na2HPO4 hardening solution. Scanning electron microscopy (SEM) was used to observe CPCmorphology. X-ray diffraction (XRD) was used to follow CPC transformation. The loading/release capacityof the CPC was studied by a bovine serum albumin-loading model. Release/retention was measured byhigh performance liquid chromatography and X-ray photoelectron spectrometry. For chemotherapeuticloading, paclitaxel (PX) was loaded onto the CPC discs by absorption. Viability of osteosarcoma U2OS andmetastatic breast cancer MDA-MB-231 cells was measured by an AlamarBlue assay. Results of SEM andXRD showed changes in CPC due to its transformation. The loading model indicated a high retentionbehavior by the CPC composition. Cell viability tests indicated a PX minimal lethal dose of 90 mg/ml. PXreleased from CPC remained active to influence cell viability. In conclusion, this study demonstrated thatCPC is a feasible delivery vector for chemotherapeutic agents.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Breast, prostate and myeloma cancers metastasize to bone ina 70e80% frequency [1e3] The bone marrow with rich sinusoidalvasculature and wide openings provides a beneficial environmentfor metastasized cancer cells. This gives metastasized cancer cellsthe opportunity to adhere and migrate into the cancellous bone[4e6]. Such metastasized cancer is mostly treated with systemicchemotherapy and/or radiation therapy. If the tumor causesa pathologic fracture, surgical removal and fixation with adjuvantchemotherapy and radiotherapy is performed [7]. Once the tumorhas developed bone metastases, the disease in most cases isincurable. However, surgical resection of these metastases andstabilization of pathologic fractures are generally considereda valuable palliative treatment [8]. The adjuvant chemotherapy iseffective to slow down tumor growth and metastasis outgrowth.

ax: þ31 (0) 243614657.).

All rights reserved.

Classical chemotherapy consists of cyclophosphamide, metho-trexate and 5-fluorouracil, while newer therapies include anthra-cyclines and texanes like epirubicin, doxorubicin and paclitaxel [7].A further, recent approach to lessen pain and disease progression isto inhibit osteoclast recruitment with bisphosphonates [9].However, there is a risk that combining such therapy with surgicalremoval of the bone metastases will weaken the bone further.Therefore, cavities need to be filled afterwards in order to maintainphysical strength of the bone and prevent or stabilize fractures [10].Bone filling materials offer the further opportunity to administerthe drug treatment locally, thus, improving effectiveness andtoleration of high doses [11]. Moreover, when treating a patient inpalliative care, the physical load would decrease significantly ifsecondary metastases could partially be addressed locally insteadof systemically.

Several studies report the addition of methotrexate to bonefillers, usually poly(methyl methacrylate) (PMMA) bone cement[12e14]. However, PMMA can only create a burst release. Further-more the temperatures involved in the polymerization exceed60 �C, which potentially is harmful for the surrounding tissues,

M.A. Lopez-Heredia et al. / Biomaterials 32 (2011) 5411e54165412

especially in poor healers such as treated cancer patients. Calciumphosphate cements (CPC) on the other hand are biocompatible andare well used for bone filling applications and minimally-invasivesurgical procedures and present advantages compared to non-injectable/shapeable CaP materials [15]. Due to their chemicaland crystalline affinity to bone tissue, CPC can be resorbed andreplaced with bone tissue in vivo [16e18]. In addition, injectableCPC can be molded to the defect shape to fill, are applied at roomtemperature and set and harden in situ by a dissolution-precipitation mechanism. The inferior mechanical properties,compared to PMMA cements, are not of great importance for thisapplication in cancellous bone as this is usually used in combina-tion with metal implants to secure mechanical stability. Therefore,the objective of this study is to develop a local drug delivery modelby using calcium phosphate cement as a carrier for chemothera-peutic agents. We hypothesize that CPC can be used as a vector todeliver these agents locally and that can be effectively directed toaffect cancerous cells.

2. Materials and methods

2.1. CPCs

The CPC powders used consisted of 85 wt% alpha-tricalcium phosphate (a-TCP;CAM Bioceramics BV, The Netherlands), 10 wt% dicalcium phosphate anhydrous(DCPA; JT Baker Chemical Co, USA) and 5 wt% precipitated hydroxyapatite (pHA;Merck, Germany). CPC samples were obtained by mixing the CPC powder with a 2%Na2HPO4 (Merck, Germany) solution in a liquid-to-powder ratio of 0.4. Briefly, 1 g ofCPC powderwas placed in a exit-closed 2ml syringe. Then, 0.4 ml of the 2% Na2HPO4

solution was added to the CPC powder and the piston was placed back into thesyringe. The syringe was placed in a mixing apparatus (Silamat� Vivadent,Liechtenstein) and it was mixed for 20 s. After mixing, a paste was obtained, whichwas injected into PTFE molds of 8 mm in diameter by 2.7 mm in height in order toobtain disc-shape samples. After molding, samples were dried at 37 �C overnightand then samples were gently removed from the molds. Scanning electronmicroscopy (SEM; JEOL 6310, JEOL, Japan) was used to observe the CPC powders andCPC morphology. X-Ray diffraction (XRD; PW3710, Philips, The Netherlands) wasused to analyze the compounds present at the CPC before and after sample prepa-ration, in order to follow its transformation.

2.2. Bovine serum albumin-loading model

In order to study the drug delivery capacity of the CPC formulation first a bovineserum albumin-loading (BSA, SigmaeAldrich, The Netherlands) model was used.BSA loaded onto CPC discs (n ¼ 3) with 0.4, 1 and 2 mg of BSA in a 100 ml drop ofdemineralized water. The solutions were allowed to absorb for 30 min and then theCPC discs were freeze-dried. The discs were placed in flat-bottomed glass flasks with3ml of Ringer’s solution for 24 h. Discs without BSA in contact with Ringer’s solutionwere used as controls. CPC discs were analyzed after loading and after a 24-himmersion period. Solutions were recovered and BSA released was measured byhigh performance liquid chromatography (HPLC; Hitachi Corp, Japan). HPLC wasperformed using a Hypersil� Gold column coupled with a guard cartridge (ThermoScientific Interscience, The Netherlands). The mobile phase consisted of acetonitrileand double distilled water (with 0.1% of formic acid). The method used was previ-ously described by Habraken et al. [19]. Briefly, a constant 40% acetonitrile with 60%of deionized water was held during 15 min for reading at a flow rate of 0.4 ml/minwith a UV detection wavelength of 280 nm. In addition, to assess for the remainingBSA, the composition of the surface was analyzed by X-Ray photoelectron spec-troscopy (XPS; Kratos AXIS Ultra, Kratos, Japan) using monochromatic Al Ka X-raysource operating at 150 W and a delay-line detector. Three discs (n ¼ 3) wereanalyzed for each loading concentration and time point (0 and 24 h). Spectra wererecorded at a 2 � 10�7 mbar background pressure. Percentages of Ca, P and Nwere quantified from a survey scan taken at 160 eV pass energy, using CasaXPSsoftware package (Casa Software Ltd, United Kingdom) and applying Kratos sensi-tivity factors. Ca and P are related to the free CaP surface, whereas the percentage ofN indicated the surface area covered with the BSA protein [20].

2.3. Paclitaxel preparation, loading and release

Paclitaxel (PX, SigmaeAldrich, The Netherlands) mother solution was preparedby dissolving at 9 mg/ml in DMSO. This solutionwas subsequently diluted in DMEM(Invitrogen, The Netherlands) cell culture medium to final concentrations of 900, 90,9, 0.9, 0.09 and 0.009 mg/ml. DMEM medium without PX was used as a control. PXwas loaded in a drop of 100 ml DMEM solution on the top of the CPC discs. The dropwas allowed to adsorb for 30 min and then the CPC discs were freeze-dried. To study

the release, the discs were placed in 24-well flat-bottomed plates with 3 ml ofDMEM for 6 and 24 h. CPC discs without PX-DMEM culture medium were used ascontrols. The concentrations on the PX solutions and the released amount of PX (6and 24 h) from the loaded CPC samples were evaluated by HPLC. Three samples percondition/solutionwere analyzed. For sample a 10 ml aliquot sample was taken threetimes to do the measurements. The mobile phase consisted of acetonitrile anddouble distilled water. The method used was described by Kim et al. [21], withmobile phases of 34% acetonitrile in deionized water for 5 min and then lineargradients to 58% acetonitrile in 16 min to 70% acetonitrile in 2 min to 34% aceto-nitrile in 4 min and finally, 34% acetonitrile held for 5 min. The flow rate was of1.0 ml/min and UV detection was performed at 227 nm.

2.4. Cell culture

Osteosarcoma U2OS cells (OST) and metastatic breast cancer cells MDA-MB-231(MDA) were cultured in DMEM. For the OST cell line DMEM contained 100 U/ml of penicillin G, 100 mg/ml of streptomycin, 10% fetal calf serum and 1 mM of L-glutamine. In the case of the MDA cells, the medium additionally contained 1 mM

sodium pyruvate and 4 mM of L-glutamine. Cell cultures were incubated at 37 �C ina 5% CO2 atmosphere. Cells were seeded in 24-well flat-bottomed plates at densitiesof 50,000 cells/well.

2.5. Cell viability

Cell viability was tested for the PX solutions as well as for the recovered solu-tions from the PX-loaded CPC and their respective controls. In the first case, cellviability from the solutions was used to study the dose efficiency and was measuredafter 15 and 39 h of interaction with PX. The cell viability measurements for the PXreleased from the CPC were performed as follows. A drop of 100 ml DMEM solutioncontaining the required amount of PX was deposited on the top of the CPC discs. Thedropwas allowed to adsorb for 30min and then the CPC discs were freeze-dried. Thediscs were then placed in 24-well flat-bottomed plates to interact with 3 ml ofDMEM for 6 and 24 h. After these release times, the solutions were recovered,sterile-filtered using a 0.22 mm cellulose acetate syringe filter (Whatman GmbH,Germany) and put in contact with the cells in the 24-well plates. After 24 h of cellcontact cell viability readings were performed using an AlamarBlue assay (BioSourceEurope, Belgium). Again, discs without PX in contact with DMEM culture mediumswere used as controls. All cell viability tests were made in triplicate (n ¼ 3).

3. Results

3.1. CPCs

The morphology of the CPC powders and hardened CPC aredisplayed in Fig. 1. CPC powders presented a morphology ofrounded a-TCP and pHA particles, whereas particles of DCPA werelonger and had remained a prismatic shape after the ball milling ofthe CaP phases. CPC, i.e. after mixing with the liquid phase andhardening, presented a more solid morphology. Still, evidentmicroporosity in the CPC structure was visible. The rounded a-TCPparticles of the CPC powders changed in shape mostly, while theprismatic DCPA particles were still distinguishable embeddedwithin the CPC morphology. XRD patterns (Fig. 2) of CPC powdersand CPCs show that the mixing and hardening had modified theintensity of some CaP phases peaks. After injection and setting ofthe CPC, the main components of the CPC are similar to the CaPphases of the starting powders (Fig. 2a), as discernable bycomparing both XRD patterns. However, after immersion in waterysolutions, the CPC transforms almost completely to an apatite typeof material creating awide region in the spectrum between 30� and32� (Fig. 2c).

3.2. Loading model

The loading model showed the behavior of the CPC formulationwith proteins or drugs. HPLC measurements proved that theamount of BSA released from the loaded CPC was small, in otherwords, indicating a high retention from the CPC for this type ofcompound. For the 0.4, 1.0 and 2.0 mg of BSA initially loaded, thereleased amount was 0.141 � 0.073, 0.113 � 0.083 and0.254 � 0.011 mg/ml, respectively. XPS analysis of the retention on

Fig. 1. Scanning electron microscopy images of (a, b) CPC powders and (c, d) hardened CPCs.

M.A. Lopez-Heredia et al. / Biomaterials 32 (2011) 5411e5416 5413

the surface (Fig. 3) corroborated the release results found by HPLC.Namely, for all concentrations tested the signal for N drops between0 and 24 h. While most of the signal remains after 24 h indicatingBSA retention. In the case of the control (CPC) there is a slightincrease on the N peak after 24 h of immersion but this is insig-nificant compared to the BSA-loaded CPC. The release is alsomirrored in the increasing signals for Ca and P, indicating the freenon-covered CPC material surface. Further, XPS proved that addinghigher amounts of BSA resulted in an increased presence of BSA.The difference between the N% at 0 and 24 h for all loaded amountsis maintained around 2%.

Fig. 2. X-Ray diffraction spectra of (a) CPC powder, (b) hardened CPC and (c) CPC aftera 3-day immersion in a watery solution.

3.3. PX release from CPC

The HPLC results from the PX release from the CPC confirmedthe observed behavior above. The amounts released were smallcompared to the initial loaded amount. After 6 h the releasedamount was evaluated to be around 14.31E-5� 2.11E-5 mg/ml. After24 h this approximately doubled to 26.40E-5 � 2.50E-5 mg/ml werereleased.

3.4. Cell viability

Fig. 4 shows the effectiveness of different PX doses in solution.For an initial cell density of 50,000 cells/well, the cells in solutionswith the lowest PX dose � 0.009 mg/ml � behaved similar to thosein solutions without PX (PX0). The highest dose of 900 mg/ml wasthemost effective for both cell lines. After 15 hMDA cells decreasedby 25% of the initial seeding density and after 39 h by 75% reachingthe medial lethal dose (LD50). In general, OST cells were well moresensitive to this cytostatic than MDA cells. The lowest concentra-tion effective was found to be 90 mg/ml as for this concentration,after 39 h the average of OST cells was found to be below the 50%level. Therefore, this concentration was used for the subsequentstudies loading PX on the CPC discs.

The results for PX released from CPC are shown in Fig. 5. As withthe dose effectiveness study, PX affected OST cell viability toa greater extent than for the MDA cells. OST cell number decreasedslower for the PX released medium compared to the PX solutions.The recovered PX solutions did not reach a LD50 for the OST cells,but still got closer to this value. For the MDA cells the PX did notresult in decreased viability but did arrest cell proliferationcompletely. The control solutions from the unloaded CPC alloweda high increment on the cell density for both release times.

Fig. 3. X-Ray photoelectron spectroscopy analysis for the percentage of phosphate (P), calcium (Ca) and nitrogen (N). D ¼ difference between the amounts before and after immersion.

M.A. Lopez-Heredia et al. / Biomaterials 32 (2011) 5411e54165414

4. Discussion

This study aimed to develop a local drug delivery model usingCPC as a carrier for chemotherapeutic agents. Although in this firststudy, for proof-of-concept, a pre-set cementwas used, it is the finalaim to produce an injectable CPC as a treatment option for bonemetastases. The cementwas loadedwith PX,which asmentioned, isa pharmaceutic used in recent chemotherapeutic approaches [7].First, a loading model using BSA confirmed that release could beachieved and loading-release properties can be optimized byloading concentration. Similar to the loadingmodel, PX release from

Fig. 4. Cell viability of OST and MDA cells after contact with diffe

the CPCswas lowand gradual, but proved effective againstMDA andOST cells. Thus, the hypothesis if loading the CPC with chemother-apeutic agents could be used as a delivery vector and be effective toaffect the cancerous cells, can be answered positively.

When regarding the study design, first it should be noted thatweinitially started to work with BSA as a model protein because of theevident toxicity of the PX. Technically, for BSA, release (by HPLC) aswell as retention (by XPS) could be regarded. For the PX only releasewas determined as it was deemed not safe to work in XPS. Still, PXcan be bonded to albumin as a delivery agent [22]. Hence, theinteraction CPC-BSA also is relevant. Furthermore, a formulation of

rent PX concentrations in solution (mg/ml) for 15 and 39 h.

Fig. 5. Cell viability of OST and MDA cells after contact with release medium from the PX-loaded CPC and from CPC controls after 6 and 24 h.

M.A. Lopez-Heredia et al. / Biomaterials 32 (2011) 5411e5416 5415

85% a-TCP,10% DCPA and 5% pHAwas chosen. The reasons for this isthat mixing of CaP compounds in combinationwith the subsequentprocess creates an injectable CPC that situates itself between therequirements for clinical applications, i.e. setting time, cohesion,reactivity and mechanical properties [15,23,24]. The CPC used herehardens by a dissolution-precipitation mechanism [15,23]. Whilesetting, itwill create a network ofmicrochannels andmicroporosity.Such a structure can create a capillarity effect when loading the PX,which in term can explainwhy loaded drugs or proteins are releasedin a relatively prolongedway. In addition, a-TCPwhichwas themainconstituent of the CPC, presented a fast transformationbyhydrolysisinto an apatite CaP type. The crystal growth associated to thistransformationmayalso account for the entrapmentof loadeddrugsor proteins within the cement [25]. Still, the initial amounts of PXreleased, were effective in reducing the OST and MDA cell densityand arresting cell growth. Full release can be expected duringdegradation of the cement, ensuring a prolonged local deliveryeffective against the formation of secondary metastases. Thiseffectiveness could be increased by increasing the PX-loadedamount, but still the effect on healthy cells will be then needed tobe evaluated in terms of overall biocompatibility.

Another technical remark should bemade on the solvent used todissolve the PX, i.e. DMSO. Considering that DMSO was present inthe culture medium when testing solutions with PX a minor effectof DMSO cannot be excluded. It is interesting that larger doses ofDMSO are potentially nocive [26,27]. Nevertheless, it is used inalternative approaches called DMSO potentiation therapy [28].DMSO makes cell membranes more permeable and could aid thechemotherapeutic agent to target cancer cells [28,29]. After hard-ening, CPC will continue to release Ca2þ ions. The use of DMSOcombined with the CPC may be creating a complexation of the PXwith the Ca2þ which increases the effectiveness of PX probably bycreating a “Trojan” effect [30e33], in other words an assistedtransport into the cell, thus effectively needing lower doses. Allsuch assumptions can only be tested in proper model systems.

When comparing our results to literature, no injectable degrad-able CPC composition exists for PX. Besides the PMMA cement, somestudies have been performed with CaP or CaP composites preforms.Previous researchers [34,35] studied the release of methotrexate(MTX) from hydroxyapatite blocks. However such materials are not

shapeable or injectable. When using loading amounts of0.63e2.38mgper 1 cm3blocks, the releaseofMTX from thismaterialafter 12 days maintained a mean concentration of 0.1e1.0 mg/ml.These concentrations were found to be efficient against tumor thecells. Lebugle et al. [36] studied the release of MTX from solid, non-shapeable CaP materials in vitro and in a rabbit femoral condylemodel. Blood tests of rabbits proved thatMTXwas below toxic levelsand after 24 h 20% of the initial loaded MTX was released from theCaP material. In another investigation, Abe et al. [37] evaluated theefficiency of loaded PX on hydroxyapatite-alginate composite beads.Their study used bonemetastasis sites in a ratmodel at the vertebralcolumn. The released PX slowed the paralysis linked to bonemetastases. In addition, the rate of survival increased by 150%.Although theusedbeadswerenot injectable, theyconcluded that thelocal delivery of chemotherapeutic agents gave better resultscompared to a systemic delivery, allowing the use of lower doses,whichmaycause less undesired secondaryeffects. As followupsteps,also our cement should be tested in a similarmodel, preferably as aninjectable to prove clinical feasibility.

Finally, the question whether the release effectiveness can beimproved remains. To overcome this, injectable polymer-ceramicscomposites could be considered in order to reduce the retentionof the CaP material. By adjusting the polymer properties stilla suitable material for bone filling applications can be maintained.Nevertheless, we have proven that it is possible to locally deliver inan effective way chemotherapeutic agents using a CPC as a deliveryvector and that low doses are needed when using this approach.

5. Conclusion

This study demonstrates that calcium phosphate cement isa feasible delivery vector for chemotherapeutic agents. Furthermaterial optimization and in vivo validation are required to achievea clinically applicable product.

Acknowledgments

The authors gratefully acknowledge the support of the TeRMSmartMix Program of the NetherlandsMinistry of Economic Affairsand the NetherlandsMinistry of Education, Culture and Science.We

M.A. Lopez-Heredia et al. / Biomaterials 32 (2011) 5411e54165416

thank Dr. Katarina Wolf, from the NCMLS-Cell Biology department,for providing the MDA-MB-231 cells.

References

[1] Lee RJ, Saylor PJ, Smith MR. Treatment and prevention of bone complicationsfrom prostate cancer. Bone 2011;48(1):88e95.

[2] Lipton A. Implications of bone metastases and the benefits of bone-targetedtherapy. Semin Oncol 2010;37(Suppl. 2):S15e29.

[3] Sterling JA, Edwards JR, Martin TJ, Mundy GR. Advances in the biology of bonemetastasis: how the skeleton affects tumor behavior. Bone 2011;48(1):6e15.

[4] Kakonen SM, Mundy GR. Mechanisms of osteolytic bone metastases in breastcarcinoma. Cancer 2003;97(3):834e9.

[5] Mastro AM, Gay CV, Welch DR. The skeleton as a unique environment forbreast cancer cells. Clin Exp Metastasis 2003;20(3):275e84.

[6] Mundy GR. Metastasis to bone: causes, consequences and therapeuticopportunities. Nat Rev Cancer 2002;2(8):584e93.

[7] Guarneri V, Conte PF. The curability of breast cancer and the treatment ofadvanced disease. Eur J Nucl Med Mol Imaging 2004;31:S149e61.

[8] Keller ET, Dai JL, Escara-Wilke J, Hall CL, Ignatoski K, Taichman RS, et al. Newtrends in the treatment of bone metastasis. J Cell Biochem 2007;102(5):1095e102.

[9] Lipton A, Berenson JR, Body JJ, Boyce BF, Bruland OS, Carducci MA, et al.Advances in treating metastatic bone cancer: summary statement for the firstCambridge Conference. Clin Cancer Res 2006;12(20):6209Se12S.

[10] ProloDJ, Rodrigo JJ. Contemporarybonegraft physiologyand surgery. ClinOrthopRelat Res 1985;(200):322e42.

[11] Verron E, Khairoun I, Guicheux J, Bouler JM. Calcium phosphate biomaterialsas bone drug delivery systems: a review. Drug Discov Today 2010;15(13e14):547e52.

[12] Maccauro G, Cittadini A, Casarci M, Muratori F, De Angelis D, Piconi C, et al.Methotrexate-added acrylic cement: biological and physical properties.J Mater Sci-Mater Med 2007;18(5):839e44.

[13] Wang HM, Galasko CSB, Crank S, Oliver G, Ward CA. Methotrexate loadedacrylic cement in the management of skeletal metastases - Biomechanical,biological, and systemic effect. Clin Orthop Relat Res 1995;(312):173e86.

[14] Draenert FG, Draenert K. Methotrexate-loaded polymethylmethacrylate bonecement for local bone metastasis therapy: pilot animal study in the rabbitpatellar groove. Chemotherapy 2008;54(5):412e6.

[15] Bohner M, Gbureck U, Barralet JE. Technological issues for the development ofmore efficient calcium phosphate bone cements: a critical assessment.Biomaterials 2005;26(33):6423e9.

[16] BohnerM. Silicon-substituted calciumphosphates - A critical view. Biomaterials2009;30(32):6403e6.

[17] Weiss P, Obadia L, Magne D, Bourges X, Rau C, Weitkamp T, et al. SynchrotronX-ray microtomography (on a micron scale) provides three-dimensionalimaging representation of bone ingrowth in calcium phosphate biomaterials.Biomaterials 2003;24(25):4591e601.

[18] Yuan H, Li Y, de Bruijn JD, de Groot K, Zhang X. Tissue responses of calciumphosphate cement: a study in dogs. Biomaterials 2000;21(12):1283e90.

[19] Habraken WJ, Wolke JG, Mikos AG, Jansen JA. PLGA microsphere/calciumphosphate cement composites for tissue engineering: in vitro release anddegradation characteristics. J Biomater Sci Polym Ed 2008;19(9):1171e88.

[20] Feng B, Chen J, Zhang X. Interaction of calcium and phosphate in apatite coatingon titanium with serum albumin. Biomaterials 2002;23(12):2499e507.

[21] Kim HW, Knowles JC, Kim HE. Hydroxyapatite and gelatin composite foamsprocessed via novel freeze-drying and crosslinking for use as temporary hardtissue scaffolds. J Biomed Mater Res A 2005;72(2):136e45.

[22] Zhou Q, Ching AK, Leung WK, Szeto CY, Ho SM, Chan PK, et al. Noveltherapeutic potential in targeting microtubules by nanoparticle albumin-bound paclitaxel in hepatocellular carcinoma. Int J Oncol 2011;38(3):721e31.

[23] Lopez-Heredia MA, Bohner M, Zhou W, Winnubst AJA, Wolke JGC, Jansen JA.The effect of ball milling grinding pathways on the bulk and reactivityproperties of calcium phosphate cements. J Biomed Mater Res B Appl Bio-mater, in press.

[24] Camiré CL, Gbureck U, Hirsiger W, Bohner M. Correlating crystallinity and reac-tivity in an [alpha]-tricalcium phosphate. Biomaterials 2005;26(16):2787e94.

[25] Ginebra MP, Delgado JA, Harr I, Almirall A, Del Valle S, Planell JA. Factorsaffecting the structure and properties of an injectable self-setting calciumphosphate foam. J Biomed Mater Res A 2007;80(2):351e61.

[26] Noel PR, Barnett KC, Davies RE, Jolly DW, Leahy JS, Mawdesley-Thomas LE,et al. The toxicity of dimethyl sulphoxide (DMSO) for the dog, pig, rat andrabbit. Toxicology 1975;3(2):143e69.

[27] Rubin LF, Barnett KC. Ocular effects of oral and dermal application of dimethylsulfoxide in animals. Ann NY Acad Sci 1967;141(1):333e45.

[28] Warren J, Sacksteder MR, Jarosz H, Wasserman B, Andreotti PE. Potentiation ofantineoplastic compounds by oral dimethyl sulfoxide in tumor-bearing rats.Ann NY Acad Sci 1975;243:194e208.

[29] Dorer RK, Zhong S, Tallarico JA, Wong WH, Mitchison TJ, Murray AW. A small-molecule inhibitor of Mps1 blocks the spindle-checkpoint response to a lackof tension on mitotic chromosomes. Curr Biol 2005;15(11):1070e6.

[30] Liu T, Tang A, Zhang G, Chen Y, Zhang J, Peng S, et al. Calcium phosphatenanoparticles as a novel nonviral vector for efficient transfection of DNA incancer gene therapy. Cancer Biother Radiopharm 2005;20(2):141e9.

[31] Allen TM, Cheng WW, Hare JI, Laginha KM. Pharmacokinetics and phar-macodynamics of lipidic nano-particles in cancer. Anti-Cancer Agent Med Chem2006;6(6):513e23.

[32] Roger M, Clavreul A, Venier-Julienne M-C, Passirani C, Montero-Menei C,Menei P. The potential of combinations of drug-loaded nanoparticle systemsand adult stem cells for glioma therapy. Biomaterials 2011;32(8):2106e16.

[33] Yao H-J, Ju R-J, Wang X-X, Zhang Y, Li R-J, Yu Y, et al. The antitumor efficacy offunctional paclitaxel nanomicelles in treating resistant breast cancers by oraldelivery. Biomaterials 2011;32(12):3285e302.

[34] Itokazu M, Esaki M, Yamamoto K, Tanemori T, Kasai T. Local drug deliverysystem using ceramics: vacuummethod for impregnating a chemotherapeuticagent into a porous hydroxyapatite block. J Mater Sci Mater Med 1999;10(4):249e52.

[35] Itokazu M, Sugiyama T, Ohno T, Wada E, Katagiri Y. Development of porousapatite ceramic for local delivery of chemotherapeutic agents. J Biomed MaterRes 1998;39(4):536e8.

[36] Lebugle A, Rodrigues A, Bonnevialle P, Voigt JJ, Canal P, Rodriguez F. Study ofimplantable calcium phosphate systems for the slow release of methotrexate.Biomaterials 2002;23(16):3517e22.

[37] Abe T, Sakane M, Ikoma T, Kobayashi M, Nakamura S, Ochiai N. Intraosseousdelivery of paclitaxel-loaded hydroxyapatitealginate composite beads delayingparalysis causedbymetastatic spine cancer in rats. JNeurosurg Spine2008;9(5):502e10.